<?xml version="1.0" encoding="utf-8" standalone="yes" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>SKKU IRIS Lab</title>
    <link>https://iris-lab.skku.edu/</link>
      <atom:link href="https://iris-lab.skku.edu/index.xml" rel="self" type="application/rss+xml" />
    <description>SKKU IRIS Lab</description>
    <generator>Wowchemy (https://wowchemy.com)</generator><language>en-us</language><lastBuildDate>Tue, 18 Aug 2026 00:00:00 +0000</lastBuildDate>
    <image>
      <url>https://iris-lab.skku.edu/media/icon_hub30d19cd1ec39a1dd8652017ce0a280f_82173_512x512_fill_lanczos_center_3.png</url>
      <title>SKKU IRIS Lab</title>
      <link>https://iris-lab.skku.edu/</link>
    </image>
    
    <item>
      <title>Kang Eun Jeon Appointed Assistant Professor at DGIST</title>
      <link>https://iris-lab.skku.edu/post/faculty_kangeun_jeon/</link>
      <pubDate>Tue, 18 Aug 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/faculty_kangeun_jeon/</guid>
      <description>&lt;div style=&#34;display: flex; gap: 0.5rem; margin-bottom: 1.5rem;&#34;&gt;
  &lt;img src=&#34;kejeon.jpg&#34; alt=&#34;Prof. Kang Eun Jeon&#34; style=&#34;width: calc(50% - 0.25rem); aspect-ratio: 1 / 1; object-fit: cover; object-position: 50% 30%;&#34;&gt;
  &lt;img src=&#34;dgist.jpg&#34; alt=&#34;DGIST Campus&#34; style=&#34;width: calc(50% - 0.25rem); aspect-ratio: 1 / 1; object-fit: cover;&#34;&gt;
&lt;/div&gt;
&lt;p&gt;IRIS Lab에서 박사후연구원으로 함께했던 전강은 박사님이 2026년 8월 18일 DGIST 전기전자컴퓨터공학과 조교수로 부임하셨습니다.&lt;/p&gt;
&lt;p&gt;전강은 박사님은 홍콩과학기술대학교(HKUST)에서 전자공학 학사 및 전자컴퓨터공학 박사 학위를 취득한 후, 2022년 12월부터 2025년 8월까지 IRIS Lab에서 박사후연구원으로 재직하였습니다. 이후 2025년 9월부터 KAIST AI대학원에서 박사후연구원으로 연구를 이어왔습니다.&lt;/p&gt;
&lt;p&gt;주요 연구 분야는 Efficient AI와 Processing-in-Memory (PIM)로, 인메모리 컴퓨팅 하드웨어의 특성을 고려한 Quantization, Pruning, Weight Mapping 등 SW-HW Co-design을 통해 AI 모델의 연산 및 메모리 효율을 높이는 연구를 수행해 왔습니다. 최근에는 이러한 연구를 생성형 AI로 확장하여, 효율적인 추론을 위한 Flexible &amp;amp; Adaptive Model Compression 연구를 진행하고 있습니다.&lt;/p&gt;
&lt;p&gt;그동안 ICCV, NeurIPS, ICCAD, DATE, ISLPED 등 주요 국제학회 및 저널에 다수의 연구 성과를 발표하였으며, 삼성전자 종합기술원(SAIT)과의 인메모리 컴퓨팅 공동연구 및 산업체와의 협력을 통해 연구 성과의 실제 시스템 적용에도 힘써 왔습니다.&lt;/p&gt;
&lt;p&gt;오랜 기간 IRIS Lab의 구성원으로 함께 연구했던 전강은 박사님의 교수 임용을 진심으로 축하드리며, 앞으로 DGIST에서 펼쳐갈 연구와 교육 활동을 응원합니다.&lt;/p&gt;
&lt;p&gt;Kang Eun Jeon, a former postdoctoral researcher at IRIS Lab, joined the Department of Electrical Engineering and Computer Science at DGIST as an Assistant Professor on August 18, 2026.&lt;/p&gt;
&lt;p&gt;Kang Eun received his B.S. in Electronic Engineering and Ph.D. in Electronic and Computer Engineering from the Hong Kong University of Science and Technology (HKUST). He subsequently joined IRIS Lab as a postdoctoral researcher from December 2022 to August 2025, and continued his research at the KAIST Graduate School of AI from September 2025.&lt;/p&gt;
&lt;p&gt;His research primarily focuses on Efficient AI and Processing-in-Memory (PIM), with particular emphasis on SW-HW co-design techniques including quantization, pruning, and weight mapping that account for the characteristics of in-memory computing hardware. More recently, he has expanded his research toward Flexible &amp;amp; Adaptive Model Compression for efficient generative AI inference.&lt;/p&gt;
&lt;p&gt;His work has been published in leading journals and conferences including ICCV, NeurIPS, ICCAD, DATE, and ISLPED. He has also collaborated with Samsung Advanced Institute of Technology (SAIT) on in-memory computing and worked with industry partners to translate efficient AI research into practical systems.&lt;/p&gt;
&lt;p&gt;Congratulations to Kang Eun on his faculty appointment at DGIST. We wish him continued success in his research and teaching endeavors!&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Visits Purdue University for Research Exchange</title>
      <link>https://iris-lab.skku.edu/post/visiting_purdue/</link>
      <pubDate>Mon, 03 Aug 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/visiting_purdue/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님과 소재현, 황찬욱 박사과정 학생은 2026년 8월 3일 미국 Purdue University의 Younghyun Kim 교수님이 이끄는 NEIS Lab을 방문하여 세미나 및 연구 교류를 진행하였습니다.&lt;/p&gt;
&lt;p&gt;이번 방문은 효율적인 엣지 컴퓨팅을 위한 AI 하드웨어 설계를 중심으로 양 연구실 간 연구 교류를 확대하기 위해 마련되었습니다. 특히 algorithm–hardware co-design, embedded AI architecture, energy-efficient machine learning 등 양 연구실의 공통 관심 분야를 중심으로 연구 방향과 협력 가능성을 논의하였습니다.&lt;/p&gt;
&lt;p&gt;방문 중 고종환 교수님은 세미나를 통해 IRIS Lab의 연구 방향과 관련 연구를 소개하였습니다. 이어진 공동 미팅에서는 IRIS Lab과 NEIS Lab의 학생들이 각자의 연구를 발표하고, AI 하드웨어 및 효율적인 인공지능 시스템 설계와 관련된 연구 아이디어와 향후 연구 방향에 대해 의견을 나누었습니다.&lt;/p&gt;
&lt;p&gt;또한 고종환 교수님은 Purdue University의 Gupta 교수님과 별도의 미팅을 갖고 관련 연구 주제와 향후 협력 가능성에 대해 논의하였습니다.&lt;/p&gt;
&lt;p&gt;이번 방문을 통해 양 연구실의 연구 현황과 관심 분야를 공유하고, AI hardware 및 SW–HW co-design 분야에서 지속적인 연구 교류와 협력 가능성을 모색하는 뜻깊은 기회를 가졌습니다.&lt;/p&gt;
&lt;p&gt;Under the supervision of Prof. Jong Hwan Ko, Ph.D. students Jaehyeon So and Chanwook Hwang from the IRIS Lab visited the NEIS Lab, led by Prof. Younghyun Kim at Purdue University, on August 3, 2026, for a seminar and research exchange.&lt;/p&gt;
&lt;p&gt;The visit was organized to strengthen research exchange between the two labs in AI hardware design for efficient edge computing. In particular, discussions focused on shared research interests including algorithm–hardware co-design, embedded AI architectures, and energy-efficient machine learning.&lt;/p&gt;
&lt;p&gt;During the visit, Prof. Jong Hwan Ko gave a seminar introducing the research directions and related work of the IRIS Lab. This was followed by a joint meeting in which students from the IRIS Lab and NEIS Lab presented their ongoing research and exchanged ideas on AI hardware and efficient AI system design.&lt;/p&gt;
&lt;p&gt;Prof. Jong Hwan Ko also met separately with Prof. Gupta at Purdue University to discuss related research topics and potential directions for future collaboration.&lt;/p&gt;
&lt;p&gt;Through this visit, the two labs shared their ongoing research and common interests and explored opportunities for continued research exchange and collaboration in AI hardware and SW–HW co-design.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[J39] Multi-Centroid Hyperdimensional Computing for Compact IMC Arrays via Dimension Pruning</title>
      <link>https://iris-lab.skku.edu/publication/j39_ieee_iot_2026/</link>
      <pubDate>Sat, 01 Aug 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j39_ieee_iot_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C106] ACA-GS: Adaptive-Capacity Anchored Gaussian Splatting for Compact Dynamic Radiance Fields</title>
      <link>https://iris-lab.skku.edu/publication/c106_acm_mm_2026/</link>
      <pubDate>Tue, 07 Jul 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c106_acm_mm_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C105] R-ESC: Robustly Erasing Space Concepts via Stochastic Feature Remapping</title>
      <link>https://iris-lab.skku.edu/publication/c105_eccv_2026/</link>
      <pubDate>Mon, 29 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c105_eccv_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Wins Top Excellence Award at the Nota AI On-Device AI Optimization Challenge</title>
      <link>https://iris-lab.skku.edu/post/nota/</link>
      <pubDate>Mon, 29 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/nota/</guid>
      <description>&lt;p&gt;IRIS 연구실 소재현, 연상흠, 강도영 연구원으로 구성된 Team IRIS가 한국정보과학회 KCC 2026에서 개최된 Nota AI 온디바이스 AI 최적화 경진대회에서 최우수상을 수상했습니다.&lt;/p&gt;
&lt;p&gt;본 대회는 총 3단계로, 1단계에서는 Nota AI의 CLI 기반 온디바이스 최적화 플랫폼인 NetsPresso를 활용한 모델 최적화를, 2단계에서는 Raspberry Pi 5 환경에서의 자율 최적화를, 마지막 본선에서는 발표 평가를 진행했습니다. Team IRIS는 각 단계에서 우수한 성적을 거두며 최종 최우수상을 수상했습니다.&lt;/p&gt;
&lt;p&gt;이번 챌린지는 알고리즘과 시스템 두 측면을 모두 아우르는 과제로, 팀원들이 이론과 실무 양면에서 깊이 있는 경험을 쌓는 계기가 되었고, 실제 산업 환경에 적용 가능한 실용적 AI 개발을 향한 첫걸음이 되기를 기대합니다.&lt;/p&gt;
&lt;p&gt;Team IRIS, composed of researchers Jaehyeon So, Sangheum Yeon, and Do Yeong Kang from the IRIS Lab, has won the Top Excellence Award at the Nota AI On-Device AI Optimization Challenge, held as part of KCC 2026 by the Korea Computer Congress.&lt;/p&gt;
&lt;p&gt;The competition consisted of three stages: model optimization using NetsPresso, Nota AI&amp;rsquo;s CLI-based on-device optimization platform; autonomous optimization on Raspberry Pi 5; and a final presentation evaluation. Team IRIS demonstrated strong, consistent performance across all three stages to claim the top honor.&lt;/p&gt;
&lt;p&gt;The challenge required participants to address both algorithmic and system-level considerations simultaneously, providing the team with hands-on experience that bridged theory and practice. This achievement marks a meaningful step toward developing practical AI solutions applicable to real-world industry environments.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab at KCC &amp; IEIE 2026: Research Presentations in Jeju</title>
      <link>https://iris-lab.skku.edu/post/kcc/</link>
      <pubDate>Thu, 18 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/kcc/</guid>
      <description>&lt;p&gt;IRIS 연구실은 2026년 6월 23일(화)부터 26일(금)까지 제주에서 개최된 대한전자공학회 하계종합학술대회와 한국정보과학회 KCC 2026에 참가하며 연구실 워크샵을 진행하였습니다.&lt;/p&gt;
&lt;p&gt;이번 학술대회에서 IRIS 연구실은 대한전자공학회에서 구두 발표 1편과 포스터 6편을 발표하였습니다. 3D·4D 가우시안 스플래팅, 확산 모델의 개념 언러닝, 산업 이상 이미지 생성, Diffusion LLM 블록 분할, 신경망 경량화 등 생성 모델과 온디바이스 효율화를 아우르는 다양한 분야에서 최신 연구 성과를 선보였습니다.&lt;/p&gt;
&lt;p&gt;특히 KCC 2026에서 개최된 Nota AI 온디바이스 AI 최적화 경진대회에서는 Team IRIS(연상흠, 강도영, 소재현)가 최우수상을 수상하는 값진 성과를 거두었습니다. NetsPresso 기반 모델 최적화부터 Raspberry Pi 5 자율 최적화, 본선 발표에 이르는 전 단계에서 우수한 성적을 기록하며 이룬 결실이었습니다.&lt;/p&gt;
&lt;p&gt;학술 일정 외에도 배낚시 체험과 우도 투어 등 제주의 다양한 활동을 함께 즐기며 구성원들 간 교류와 친목을 다지는 시간을 가졌습니다. 연구에 대한 열정을 나누는 동시에 서로를 더 깊이 이해하는 계기가 되었으며, 앞으로의 연구 여정에도 큰 힘이 될 것입니다.&lt;/p&gt;
&lt;p&gt;The IRIS Lab participated in the IEIE Summer Annual Conference and KCC 2026 (Korea Computer Congress), held in Jeju from June 23 (Tue) to June 26 (Fri), 2026.&lt;/p&gt;
&lt;p&gt;At the IEIE conference, the IRIS Lab presented one oral paper and six posters. Spanning topics such as 3D and 4D Gaussian Splatting, concept unlearning in diffusion models, industrial anomaly image generation, block partitioning for diffusion LLMs, and neural network compression, the lab showcased its latest research across a range of areas bridging generative models and on-device efficiency.&lt;/p&gt;
&lt;p&gt;Most notably, at the Nota AI On-Device AI Optimization Challenge held as part of KCC 2026, Team IRIS (Sangheum Yeon, Doyeong Kang, and Jaehyeon So) earned the Top Excellence Award. The team delivered strong performance across every stage, from NetsPresso-based model optimization to autonomous optimization on Raspberry Pi 5 and the final presentation, achieving this remarkable result.&lt;/p&gt;
&lt;p&gt;Beyond the academic program, members enjoyed a variety of activities in Jeju, including a sea fishing experience and a tour of Udo Island, sharing time together and strengthening their bonds. It was an opportunity to exchange research passion while getting to know one another more deeply, and one that will surely energize the journey ahead.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Attends 2026 IEEE/JSAP Symposium on VLSI Technology and Circuits</title>
      <link>https://iris-lab.skku.edu/post/vlsi/</link>
      <pubDate>Thu, 18 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/vlsi/</guid>
      <description>&lt;p&gt;IRIS 연구실의 소재현 박사과정 학생과 황찬욱 박사과정 학생은 2026년 6월 15일부터 18일까지 미국 하와이 Hilton Hawaiian Village에서 개최된 2026 IEEE/JSAP Symposium on VLSI Technology and Circuits에 참석했습니다.&lt;/p&gt;
&lt;p&gt;이번 학회에서는 최재혁 교수님 연구실(EEMIX Lab)의 이호욱 연구원이 발표한 논문 “All-Digital Event-based Vision Sensor with Scene Adaptive Power-Saving Pixels and Three-Layer Neural Network for Object Detection”이 소개되었습니다. 본 논문은 EEMIX Lab과의 협업을 통해 수행된 연구로, 소재현 학생은 공동 제1저자, 황찬욱 학생은 제2저자로 참여했습니다.&lt;/p&gt;
&lt;p&gt;본 연구는 All-Digital Event-based Vision Sensor에 장면 적응형 전력 절감 픽셀 구조와 3-layer neural network 기반 객체 검출 기능을 결합한 시스템을 제안합니다. 이를 통해 불필요한 이벤트와 전력 소모를 줄이면서도 효율적인 객체 인식을 수행할 수 있는 가능성을 보였습니다.&lt;/p&gt;
&lt;p&gt;IEEE/JSAP Symposium on VLSI Technology and Circuits는 반도체 소자 기술 및 집적회로 설계 분야의 대표적인 국제 학회로, 전 세계 연구자들이 최신 회로 설계 및 반도체 기술 동향을 공유하는 자리입니다. 소재현 학생과 황찬욱 학생은 이번 학회 참석을 통해 최신 VLSI 회로 설계, 지능형 센서 시스템, 저전력 AI 하드웨어 분야의 연구 동향을 폭넓게 접하고, 국내외 연구자들과 교류하는 뜻깊은 시간을 가졌습니다.&lt;/p&gt;
&lt;p&gt;이번 참석은 IRIS 연구실이 수행하고 있는 이벤트 기반 비전, 인-센서 컴퓨팅, 저전력 AI 시스템 연구가 세계적인 회로 학회 무대와 연결되었다는 점에서 의미 있는 경험이 되었습니다.&lt;/p&gt;
&lt;p&gt;Jaehyeon So and Chanwook Hwang from IRIS Lab attended the 2026 IEEE/JSAP Symposium on VLSI Technology and Circuits, held from June 15 to 18, 2026, at Hilton Hawaiian Village in Hawaii, USA.&lt;/p&gt;
&lt;p&gt;At the symposium, Houwook Lee from Prof. Jaehyouk Choi’s EEMIX Lab presented the paper titled “All-Digital Event-based Vision Sensor with Scene Adaptive Power-Saving Pixels and Three-Layer Neural Network for Object Detection.” This work was conducted in collaboration with EEMIX Lab, with Jaehyeon So contributing as a co-first author and Chanwook Hwang as the second author.&lt;/p&gt;
&lt;p&gt;This paper proposes an all-digital event-based vision sensor that integrates scene-adaptive power-saving pixels with a three-layer neural network for object detection. The proposed system aims to reduce unnecessary events and power consumption while enabling efficient object recognition.&lt;/p&gt;
&lt;p&gt;The IEEE/JSAP Symposium on VLSI Technology and Circuits is one of the leading international conferences in semiconductor device technology and integrated circuit design. Through this conference, Jaehyeon So and Chanwook Hwang had the opportunity to learn about recent trends in VLSI circuit design, intelligent sensor systems, and low-power AI hardware, while engaging with researchers from academia and industry.&lt;/p&gt;
&lt;p&gt;This attendance was a meaningful experience for IRIS Lab, connecting our ongoing research in event-based vision, in-sensor computing, and energy-efficient AI systems with the global VLSI research community.&lt;/p&gt;
&lt;p&gt;Related News: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=219294&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;성균관대학교 정보통신대학, 세계 최상위 반도체 회로 학회 VLSI Symposium에 논문 6편 채택&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C104] Foundation Model-Guided RGB-to-RAW Generation with Spectral Supervision for RAW-Domain Detection</title>
      <link>https://iris-lab.skku.edu/publication/c104_avss_2026/</link>
      <pubDate>Tue, 16 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c104_avss_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C103] UICAM: A Codec-Transferable Adapter for Machine-Oriented Image Compression</title>
      <link>https://iris-lab.skku.edu/publication/c103_avss_2026/</link>
      <pubDate>Mon, 15 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c103_avss_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Prof. Younghyun Kim’s Seminar on ‘Time-Accuracy Scalable Compute for Energy-Efficient Machine Learning&#39;</title>
      <link>https://iris-lab.skku.edu/post/seminar_younghyun_kim-copy/</link>
      <pubDate>Mon, 15 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/seminar_younghyun_kim-copy/</guid>
      <description>&lt;p&gt;우리 연구실은 2026년 6월 15일, Purdue University 전자전기컴퓨터공학과의 Younghyun Kim 교수님을 모시고 &amp;ldquo;Time-Accuracy Scalable Compute for Energy-Efficient Machine Learning&amp;quot;을 주제로 한 세미나를 개최했습니다.&lt;/p&gt;
&lt;p&gt;Younghyun Kim 교수님은 Purdue University Elmore Family School of Electrical and Computer Engineering의 부교수로, Networked and Embedded Intelligent Systems Lab을 이끌고 계시며 에너지-품질 확장형 컴퓨팅, 엣지/응용/물리 AI, 사이버 물리 시스템 등을 연구하고 계십니다.&lt;/p&gt;
&lt;p&gt;이번 세미나에서는 전통적인 반도체 미세화를 넘어 에너지 효율적인 머신러닝을 실현하기 위한 방법론으로 시간-정확도 확장형 컴퓨팅(time-accuracy scalable computing)이 소개되었습니다. 근사 컴퓨팅(approximate computing)을 통해 정확도를 성능 및 에너지 효율과 맞교환함으로써, 동적 정확도 제어와 전 시스템 최적화가 머신러닝의 정확도를 유지하면서도 지연시간, 전력, 에너지를 크게 절감할 수 있음을 보여주었습니다.&lt;/p&gt;
&lt;p&gt;귀중한 강연으로 깊은 통찰을 나누어 주신 Younghyun Kim 교수님께 진심으로 감사드립니다.&lt;/p&gt;
&lt;p&gt;On June 15, 2026, our lab hosted a seminar featuring Professor Younghyun Kim from the School of Electrical and Computer Engineering at Purdue University, on the topic of &amp;ldquo;Time-Accuracy Scalable Compute for Energy-Efficient Machine Learning.&amp;rdquo;
Younghyun Kim is an Associate Professor in the Elmore Family School of Electrical and Computer Engineering at Purdue University, where he leads the Networked and Embedded Intelligent Systems Lab. His research interests include energy-quality scalable computing, edge/applied/physical AI, cyber-physical systems, and security and privacy for embedded computing systems.
In this seminar, Professor Younghyun Kim presented time-accuracy scalable computing as a path toward energy-efficient machine learning beyond traditional semiconductor scaling. He showed how approximate computing trades accuracy for improved performance and energy efficiency, demonstrating that dynamic accuracy control and full-system optimization can substantially reduce latency, power, and energy while maintaining acceptable machine-learning accuracy.
We sincerely thank Professor Younghyun Kim for sharing his valuable insights through this enlightening lecture.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Prof. Jong Hwan Ko&#39;s Research Team Wins Best Paper Award at CVPR 2026 Embedded Vision Workshop</title>
      <link>https://iris-lab.skku.edu/post/award_cvpr/</link>
      <pubDate>Wed, 03 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/award_cvpr/</guid>
      <description>&lt;p&gt;정보통신대학 고종환 교수 연구팀의 소재현 박사과정 학생이 미국 덴버(Denver)에서 개최된 CVPR(Conference on Computer Vision and Pattern Recognition) 2026 Embedded Vision Workshop에서 Best Paper Award를 수상하였다.&lt;/p&gt;
&lt;p&gt;CVPR은 컴퓨터 비전 및 인공지능 분야에서 가장 권위 있는 국제학술대회 중 하나로 평가받으며, Embedded Vision Workshop은 저전력·고효율 비전 시스템과 엣지 인공지능 기술 분야의 최신 연구 성과를 공유하는 전문 워크숍이다.&lt;/p&gt;
&lt;p&gt;수상 논문인 &amp;ldquo;EventGuard: Sparsity-Aware In-Sensor Denoising for Frame-Based Event Vision Sensors&amp;quot;는 프레임 기반 이벤트 비전 센서를 위한 인-센서 노이즈 제거 기술을 제안하였다.&lt;/p&gt;
&lt;p&gt;본 연구는 이벤트 데이터의 희소성을 활용하여 센서 내부에서 노이즈를 제거하고, 유효한 이벤트에 대해서만 연산을 수행하는 구조를 제시하였다. 이를 통해 불필요한 연산을 줄이면서도 효과적인 노이즈 제거를 가능하게 하였으며, 저전력·고효율 이벤트 처리 시스템 구현 가능성을 보여주었다.&lt;/p&gt;
&lt;p&gt;해당 기술은 차세대 저전력 인공지능 시스템 및 임베디드 비전 시스템에 활용될 수 있을 것으로 기대되며, 이번 수상은 성균관대학교 전자전기컴퓨터공학과의 이벤트 비전 및 인-센서 컴퓨팅 분야 연구 역량을 국제적으로 인정받은 성과로 평가된다.&lt;/p&gt;
&lt;p&gt;Jaehyeon So, a doctoral student in Professor Jong Hwan Ko&amp;rsquo;s research team at the College of Information and Communication Engineering, has received the Best Paper Award at the CVPR (Conference on Computer Vision and Pattern Recognition) 2026 Embedded Vision Workshop, held in Denver, USA.&lt;/p&gt;
&lt;p&gt;CVPR is recognized as one of the most prestigious international conferences in the fields of computer vision and artificial intelligence. The Embedded Vision Workshop is a specialized forum for sharing the latest research in low-power, high-efficiency vision systems and edge AI technologies.&lt;/p&gt;
&lt;p&gt;The award-winning paper, &amp;ldquo;EventGuard: Sparsity-Aware In-Sensor Denoising for Frame-Based Event Vision Sensors,&amp;rdquo; proposes an in-sensor denoising technique for frame-based event vision sensors.&lt;/p&gt;
&lt;p&gt;The research leverages the sparsity of event data to perform noise removal directly within the sensor, computing only on valid events. This approach reduces unnecessary computation while enabling effective denoising, demonstrating the feasibility of low-power, high-efficiency event processing systems.&lt;/p&gt;
&lt;p&gt;The technology is expected to find applications in next-generation low-power AI and embedded vision systems. This award is recognized as an internationally validated achievement highlighting SKKU&amp;rsquo;s Department of Electrical and Computer Engineering&amp;rsquo;s research capabilities in event vision and in-sensor computing.&lt;/p&gt;


















&lt;figure  id=&#34;figure-소재현-박사과정-학생-수상-현장&#34;&gt;
  &lt;div class=&#34;d-flex justify-content-center&#34;&gt;
    &lt;div class=&#34;w-100&#34; &gt;&lt;img alt=&#34;소재현 박사과정 학생 수상 현장&#34; srcset=&#34;
               /post/award_cvpr/award_hub7e620b237e0aaee3f533985c4b5f542_2096358_236bdc8d58230c81a1763a6baa2951f6.webp 400w,
               /post/award_cvpr/award_hub7e620b237e0aaee3f533985c4b5f542_2096358_0fd1cca03eadbdac92818931ef5e2cf1.webp 760w,
               /post/award_cvpr/award_hub7e620b237e0aaee3f533985c4b5f542_2096358_1200x1200_fit_q75_h2_lanczos_3.webp 1200w&#34;
               src=&#34;https://iris-lab.skku.edu/post/award_cvpr/award_hub7e620b237e0aaee3f533985c4b5f542_2096358_236bdc8d58230c81a1763a6baa2951f6.webp&#34;
               width=&#34;760&#34;
               height=&#34;384&#34;
               loading=&#34;lazy&#34; data-zoomable /&gt;&lt;/div&gt;
  &lt;/div&gt;&lt;figcaption&gt;
      소재현 박사과정 학생 수상 현장
    &lt;/figcaption&gt;&lt;/figure&gt;

</description>
    </item>
    
    <item>
      <title>Do Yeong Kang Admitted to Georgia Tech PhD Program</title>
      <link>https://iris-lab.skku.edu/post/phd_offer_doyeong_kang/</link>
      <pubDate>Tue, 02 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/phd_offer_doyeong_kang/</guid>
      <description>&lt;p&gt;강도영 석사과정 학생이 Georgia Tech 박사과정에 합격하였습니다. 연구조교(GRA) 및 펠로우십을 포함하여 연간 약 70,000달러 규모의 지원을 받게 되었으며, Saibal Mukhopadhyay 교수의 지도 아래 연구를 이어갈 예정입니다.&lt;/p&gt;
&lt;p&gt;석사과정 동안 인공지능 모델의 성능을 유지하면서도 연산 및 전력 소모를 최소화하는 경량화 기술과 Brain-inspired Hyperdimensional Computing 등 Efficient AI 분야에 초점을 맞추어 왔습니다. 박사 과정 동안에는 이러한 효율성 패러다임을 단일 에이전트를 넘어 협업 시스템으로 확장하여, 멀티 에이전트 시스템(Multi-Agent Systems)에서 에이전트 간 효율적인 통신(Efficient Communication)과 관련된 연구를 진행할 예정입니다.&lt;/p&gt;
&lt;p&gt;Georgia Tech의 Saibal Mukhopadhyay 교수 연구실은 하드웨어 친화적이고 효율적인 컴퓨팅 기술을 선도하고 있어, 강도영 학생이 에이전트 간 통신과 효율적 AI 연구를 심화하는 데 최적의 연구 환경이 될 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;다시 한번 강도영 학생의 박사과정 합격을 진심으로 축하하며, 앞으로의 연구 활동을 응원합니다.&lt;/p&gt;
&lt;p&gt;Doyoung Kang has been offered a fully funded PhD position at the Georgia Institute of Technology (Georgia Tech), supported by a Graduate Research Assistantship (GRA) and a fellowship amounting up to $70,000 annually. He will be advised by Professor Saibal Mukhopadhyay.&lt;/p&gt;
&lt;p&gt;During his master&amp;rsquo;s program, his research focused on Efficient AI, particularly in model lightweighting techniques to maintain performance while minimizing computational and power overhead, as well as brain-inspired hyperdimensional computing. During his PhD, Doyoung plans to expand this efficiency paradigm from a single agent to collaborative systems, investigating efficient communication between agents in Multi-Agent Systems (MAS).&lt;/p&gt;
&lt;p&gt;As Professor Saibal Mukhopadhyay’s research group at Georgia Tech leads the way in hardware-friendly and efficient computing technologies, it is expected to provide Doyoung with an optimal environment to deepen his research in agent communication and efficient AI. Go Yellow Jackets!&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Jinju Kim Admitted to UT Austin PhD Program</title>
      <link>https://iris-lab.skku.edu/post/phd_offer_jinju_kim/</link>
      <pubDate>Tue, 02 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/phd_offer_jinju_kim/</guid>
      <description>&lt;p&gt;김진주 석사과정 학생이 UT Austin 박사과정에 합격하였습니다. 연구조교(GRA) 및 펠로우십을 포함하여 연간 약 73,000달러 규모의 지원을 받게 되었으며, Atlas Wang 교수의 지도 아래 연구를 이어갈 예정입니다. 특히 IRIS Lab에서 지난 몇 년간 발전시켜 온 연구 주제인 안전하고 신뢰할 수 있는 생성형 모델을 위한 머신 언러닝(Machine Unlearning) 연구로 학문적 여정을 지속할 계획입니다.&lt;/p&gt;
&lt;p&gt;생성형 모델은 실제와 구분하기 어려운 음성, 음악, 영상 등을 생성할 수 있으며, 이는 해당 기술의 강력한 장점인 동시에 중요한 사회적·기술적 문제를 제기합니다. 예를 들어 음성 생성 모델은 특정 인물의 목소리를 매우 정교하게 재현할 수 있고, 음악 생성 모델은 학습 데이터와 유사한 결과물을 만들어낼 수 있습니다. 김진주 학생의 연구는 이러한 맥락에서 다음과 같은 질문에 주목합니다. 모델이 특정 정보를 의도적으로 잊도록 만들 수 있는가? 역으로, 특정 정보가 여전히 모델 내부에 남아 있음을 증명하거나, 온전한 삭제가 이루어졌음을 검증할 수 있는가?&lt;/p&gt;
&lt;p&gt;이 연구는 생성형 AI의 신뢰성 객관적으로 측정하고 검증할 수 있는 특성으로 정립하는 것을 목표로 합니다. UT 오스틴에서는 이러한 문제를 보다 체계적으로 탐구하며, 안전하고 신뢰할 수 있는 생성형 AI 개발에 기여할 예정입니다.&lt;/p&gt;
&lt;p&gt;UT Austin의 Atlas Wang 교수 연구실 또한 안전한 생성형 모델을 위한 정교하고 효율적인 방법론을 지속적으로 발전시켜 왔습니다. 이러한 연구 환경은 머신 언러닝과 신뢰 가능한 생성 모델 연구를 심화하는 데 매우 적합한 기반이 될 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;다시 한번 김진주 학생의 박사과정 합격을 진심으로 축하하며, 앞으로의 연구 활동을 응원합니다.&lt;/p&gt;
&lt;p&gt;Jinju Kim has been offered a fully funded PhD position at the University of Texas at Austin, supported by a Graduate Research Assistantship (GRA) and a fellowship amounting up to $73,000 annually. She will be advised by Professor Atlas Wang, continuing the research thread that has organized the last few years of her work at IRIS Lab: machine unlearning for safe, trustworthy generation.&lt;/p&gt;
&lt;p&gt;Generative models are very good at producing real-world-quality speech, music, and video — that&amp;rsquo;s the point of them, and also the problem. A speech model can reconstruct a particular voice; a music model can generate something close to what it was trained on. The question Jinju&amp;rsquo;s research takes seriously is whether these systems can let go of something on purpose, and whether that removal can be verified — even when someone is actively trying to prove the knowledge is still in there. Turning &amp;ldquo;trustworthy generation&amp;rdquo; from a slogan into a property you can measure and defend is the work she&amp;rsquo;ll be doing at Austin.&lt;/p&gt;
&lt;p&gt;She chose UT Austin because Atlas Wang&amp;rsquo;s group also builds on this problem — principled, efficient methods to manipulate and modify generative models for trustworthy. Go Longhorns!&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Jinju joins Sony AI as Summer Research Intern</title>
      <link>https://iris-lab.skku.edu/post/internship_sony/</link>
      <pubDate>Mon, 01 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/internship_sony/</guid>
      <description>&lt;p&gt;이번 여름 김진주 석사과정 연구원이 도쿄 Sony AI의 Music Foundation Model Team에 연구 인턴으로 합류하여, 아카사카 Sony Research 오피스에서 연구를 시작하게 되었습니다. 김진주 연구원은 석사과정 동안 탐구해 온 생성형 AI에서의 언러닝 기법 연구를 이어나가, 생성형 음악 모델 분야에서 AI 생성 미디어의 저작권, 기억(memorization), 언러닝이 교차하는 지점을 산업 현장에서 직접 연구하게 되었습니다.&lt;/p&gt;
&lt;p&gt;김진주 연구원은 Sony AI에서의 연구를 이렇게 표현합니다. &amp;ldquo;AI가 정확히 무엇을 하고 있는지, 그리고 무엇을 하지 말아야 하는지를 발견하는 것.&amp;rdquo; 단순한 윤리적 질문처럼 들리지만, 본질적인 해결책은 기술적 탐구에 있습니다. 생성형 모델이 실제로 어떤 지식을 학습하는지, 그 지식이 모델 내부에 어떻게 저장되는지, 그리고 어느 시점에서 그것이 저작물의 재현에 해당하는지가 연구의 핵심입니다. 특히 음악 생성 모델 분야에서 이러한 질문들은 더욱 중요한 쟁점이 됩니다. 저작권이 있는 음원으로 학습된 모델은 단순히 음악적 스타일을 흡수하는 데 그치지 않고 특정 음원을 재구성할 수도 있으며, &amp;lsquo;학습의 영향&amp;rsquo;과 &amp;lsquo;저작권 침해&amp;rsquo; 사이의 경계는 법적으로도 기술적으로도 아직 명확히 정립되지 않았습니다. 이번 여름 김진주 연구원은 언러닝 기법을 활용한 학습 데이터 귀속(Train Data Attribution) 연구에 집중하고 있습니다. AI로 생성된 음원에 기여한 학습 데이터를 추적하고, 아티스트들이 자신의 작업이 AI에 어떻게, 얼마나 반영되는지에 대한 투명성을 확보하는 것이 목표입니다.&lt;/p&gt;
&lt;p&gt;이 연구는 IRIS Lab에서의 머신 언러닝 연구와 NeurIPS 2025 AI4Music 워크숍 1저자 논문 발표로부터 이어지게 되었습니다. Sony AI에서 산업 현장의 현실적인 기술 문제를 탐구하며 사회에 기여하는 연구를 이어가길 응원합니다.&lt;/p&gt;
&lt;p&gt;Jinju Kim has joined Sony AI&amp;rsquo;s Music Foundation Model Team as a research intern this summer, based at the Sony Research Akasaka office in Tokyo. Her work sits at the intersection of generative music models and questions of AI copyright, memorization, and unlearning - territory that she has been circling for a while, now met head-on in an industry setting.&lt;/p&gt;
&lt;p&gt;She describes her research at Sony AI in a few words: &amp;ldquo;discovering what AI does, and shouldn&amp;rsquo;t do.&amp;rdquo; This sounds like an ethics talk but is really a technical one: what does a generative model actually learn, how does that knowledge persist inside it, and at what point does that constitute reproduction? Music generation makes these questions especially concrete. A model trained on copyrighted recordings doesn&amp;rsquo;t just absorb style — it can reconstruct specific material, and the line between influence and infringement is far from settled legally or technically. This summer, she is focusing on unlearning as a training data attribution method, working to ensure that the creative contributions embedded in generated outputs can be traced back to their sources, and that artists retain visibility over how their work shapes what these models produce.&lt;/p&gt;
&lt;p&gt;This is a thread Jinju has been pulling on since her work on machine unlearning at IRIS Lab, and through her first-author paper at the NeurIPS 2025 AI4Music workshop. Sony AI&amp;rsquo;s Music Foundation Model Team is happy to have her on the team: the problems are live, the stakes are real, and the gap between what current models do and what they should do is exactly what her research is designed to close.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C102] Contrastive Flow Map Matching</title>
      <link>https://iris-lab.skku.edu/publication/c102_icml_2026/</link>
      <pubDate>Sat, 30 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c102_icml_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J38] Partial-path Rectification in Diffusion Sampling</title>
      <link>https://iris-lab.skku.edu/publication/j38_ttm_2026/</link>
      <pubDate>Sat, 30 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j38_ttm_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Visits Calgary ML Lab for Research Exchange</title>
      <link>https://iris-lab.skku.edu/post/visiting_calgary/</link>
      <pubDate>Fri, 22 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/visiting_calgary/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님 인솔 하에 오영환, 강윤성 연구원은 2026년 5월 18일부터 19일까지 캐나다 캘거리 대학교(University of Calgary) Yani Ioannou 교수님의 Calgary ML Lab을 방문하여 연구 교류를 진행하였습니다.&lt;/p&gt;
&lt;p&gt;첫째 날에는 Ioannou 교수님과 개별 미팅을 통해 각 연구원의 연구에 대한 직접적인 피드백을 받았습니다. 강윤성 연구원은 디퓨전 모델 concept unlearning 연구를 공유하며, attention 기반 loss 설계, 타임스텝별 erase·anchor loss 비중을 동적 조절하는 Time-Gated Loss 기법, style erasure와 concept erasure 분리 전략, sparsity가 생성 품질에 미치는 영향 등을 논의하였습니다. 오영환 연구원은 Dynamic Sparse Training(DST)과 machine unlearning을 결합한 접근법을 소개하고, unstructured에서 structured sparsity로의 확장 방안과 classification·Stable Diffusion 모델 양쪽에의 병렬 적용 계획을 검토하였습니다.&lt;/p&gt;
&lt;p&gt;둘째 날에는 양 연구실 합동 미팅에서 소속 연구원들이 각자의 연구를 발표하였습니다. IRIS 연구실에서는 오영환 연구원이 machine unlearning 기초 개념 및 DST+Unlearning 접근법을, 강윤성 연구원이 디퓨전 모델 unlearning 기법을 각각 발표하였습니다. Calgary ML Lab에서는 Mohammed Adnan의 sparse training 안정성 보정 기법 SparseOpt (ICML 2026 accepted), Mike Lasby의 MoE one-shot 압축 기법 REAP (ICLR 2026), Yufan Feng의 VLM 대상 concept-driven backdoor 공격 연구 (ICML Workshop 2025), Tejas Pote의 data pruning과 classification bias 분석 (CPAL 2024), Abhishek Rajora의 Lottery Ticket Hypothesis 소개가 발표되었습니다.&lt;/p&gt;
&lt;p&gt;발표 후에는 sparsity 알고리즘의 하드웨어 가속 실현 가능성, unlearning 적용 후 adversarial prompt에 의한 개념 복원 취약점과 robustness 평가 기준 확립 필요성 등을 주제로 토론이 이어졌으며, 양 연구실 간 sparsity·unlearning·모델 압축 분야의 후속 협력 방향을 논의하였습니다.&lt;/p&gt;
&lt;p&gt;Under the supervision of Prof. Jong Hwan Ko, researchers Yeong Hwan Oh and Yunsung Kang from the IRIS Lab visited the Calgary ML Lab, led by Prof. Yani Ioannou at the University of Calgary, Canada, from May 18 to 19, 2026.&lt;/p&gt;
&lt;p&gt;On the first day, each researcher received direct feedback from Prof. Ioannou on their ongoing work. Yunsung Kang shared his research on concept unlearning in diffusion models, discussing attention-based loss design, the Time-Gated Loss technique for dynamically balancing erase and anchor losses across timesteps, strategies for separating style and concept erasure, and the effect of sparsity on generation quality. Yeong Hwan Oh introduced his approach combining Dynamic Sparse Training (DST) with machine unlearning and reviewed plans for extending from unstructured to structured sparsity and for parallel application to both classification and Stable Diffusion models.&lt;/p&gt;
&lt;p&gt;On the second day, a joint lab meeting featured research presentations from both groups. From the IRIS Lab, Yeong Hwan Oh presented on machine unlearning fundamentals and his DST+Unlearning approach, while Yunsung Kang presented on diffusion model unlearning. From Calgary ML Lab, presentations included Mohammed Adnan&amp;rsquo;s SparseOpt for sparse training stability (ICML 2026 accepted), Mike Lasby&amp;rsquo;s REAP for one-shot MoE compression (ICLR 2026), Yufan Feng&amp;rsquo;s concept-driven backdoor attacks on VLMs (ICML Workshop 2025), Tejas Pote&amp;rsquo;s analysis of data pruning and classification bias (CPAL 2024), and Abhishek Rajora&amp;rsquo;s introduction to the Lottery Ticket Hypothesis.&lt;/p&gt;
&lt;p&gt;After the presentations, discussions covered the feasibility of hardware acceleration for sparsity-based algorithms, the vulnerability of unlearned concepts to recovery via adversarial prompts and the need for robustness evaluation criteria, and directions for continued collaboration between the two labs in sparsity, unlearning, and model compression.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C101] HyperSPACE: Sparse-Adder-Compatible Encoding for Efficient Hyperdimensional Computing on Digital CIM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c101_islped_2026/</link>
      <pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c101_islped_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C100] Dataflow-Preserving, Overhead-Free Weight Remapping for Fault-Tolerant ReRAM-Based In-Memory Computing</title>
      <link>https://iris-lab.skku.edu/publication/c100_islped_2026/</link>
      <pubDate>Wed, 20 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c100_islped_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C99] A Neural-Feedback-Driven Event Camera for Robust and Efficient Vision Processing</title>
      <link>https://iris-lab.skku.edu/publication/c99_islped_2026/</link>
      <pubDate>Tue, 19 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c99_islped_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab 2026 Homecoming</title>
      <link>https://iris-lab.skku.edu/post/homecoming_202505/</link>
      <pubDate>Wed, 06 May 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/homecoming_202505/</guid>
      <description>&lt;p&gt;2026년 5월 6일(수), IRIS 연구실의 2026 홈커밍 행사가 진행되었습니다.&lt;/p&gt;
&lt;p&gt;홈커밍은 매년 스승의 날을 기점으로 열리는 IRIS 연구실만의 전통 행사로, 졸업생과 재학생이 한자리에 모여 서로의 근황을 나누고 교류하는 뜻깊은 자리입니다. 이번 행사는 오후 6시부터 FAFAS에서 진행되었으며, 오랜만에 만난 선후배들과 즐거운 시간을 보내며 연구실의 끈끈한 유대감을 다시 한번 확인할 수 있었습니다.&lt;/p&gt;
&lt;p&gt;On May 6, 2026, IRIS Lab held its annual Homecoming event.&lt;/p&gt;
&lt;p&gt;Homecoming is a cherished IRIS Lab tradition held each year around Teacher&amp;rsquo;s Day, bringing together alumni and current members to reconnect and share updates on each other&amp;rsquo;s lives and research. This year&amp;rsquo;s gathering took place from 6:00 PM at FAFAS (13 Seobu-ro 2105beon-gil, 8F), offering a warm opportunity to strengthen the bonds between generations of lab members.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[J37] EPS: Efficient Patch Sampling for Video Overfitting in Deep Super-Resolution Model Training</title>
      <link>https://iris-lab.skku.edu/publication/j37_ieee-tcsvt_2026/</link>
      <pubDate>Thu, 16 Apr 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j37_ieee-tcsvt_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Futsal Match vs. Prof. Kim Sejung&#39;s Lab</title>
      <link>https://iris-lab.skku.edu/post/footsal_2026/</link>
      <pubDate>Thu, 16 Apr 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/footsal_2026/</guid>
      <description>&lt;p&gt;2026년 4월 16일(목), IRIS 연구실 구성원들이 김세정 교수님 연구실과 풋살 친선 경기를 가졌습니다.&lt;/p&gt;
&lt;p&gt;오후 5시부터 7시까지 HK풋살파크 7구장에서 진행된 이번 경기는 연구실 간 교류와 친목을 다지는 즐거운 자리였습니다. 평소 연구와 공부로 바쁜 일상 속에서 함께 땀을 흘리며 활력을 충전하는 뜻깊은 시간이 되었습니다.&lt;/p&gt;
&lt;p&gt;On April 16, 2026, IRIS Lab members took part in a friendly futsal match against the lab of Prof. Kim Sejung.&lt;/p&gt;
&lt;p&gt;The match was held from 5:00 to 7:00 PM at HK Futsal Park, Court 7. It was a great opportunity to build camaraderie between the two labs and recharge with some friendly competition after long days of research.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C98] Harnessing Linguisitc Dissimilarity for Language Generalization on Unseen Low-Resource Varieties</title>
      <link>https://iris-lab.skku.edu/publication/c98_conll_2026/</link>
      <pubDate>Wed, 15 Apr 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c98_conll_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C97] RangeGuard: Efficient, Bounded Approximate Error Correction for Reliable DNNs</title>
      <link>https://iris-lab.skku.edu/publication/c97_isca_2026/</link>
      <pubDate>Fri, 27 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c97_isca_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C96] EventGuard: Sparsity-Aware In-Sensor Denoising for Frame-Based Event Vision Sensors</title>
      <link>https://iris-lab.skku.edu/publication/c96_cvpr_workshop_2026/</link>
      <pubDate>Thu, 26 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c96_cvpr_workshop_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C95] All-Digital Event-based Vision Sensor with Scene Adaptive Power-Saving Pixels and Three-Layer Neural Network for Object Detection</title>
      <link>https://iris-lab.skku.edu/publication/c95_vlsi_2026/</link>
      <pubDate>Wed, 25 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c95_vlsi_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Visits University of Arizona for Research Exchange</title>
      <link>https://iris-lab.skku.edu/post/visiting_arizona/</link>
      <pubDate>Mon, 09 Mar 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/visiting_arizona/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님과 박찬웅 박사과정 학생, 한석호 학부연구생은 2026년 3월 9일 미국 University of Arizona의 Huanrui Yang 교수님 연구실을 방문하여 세미나 및 연구 교류를 진행하였습니다.&lt;/p&gt;
&lt;p&gt;이번 방문은 효율적인 인공지능과 AI 모델 경량화를 중심으로 연구 교류를 확대하기 위해 마련되었습니다. 특히 model quantization, efficient AI inference, hardware-aware optimization, HW–SW co-design 등 공통 관심 분야를 중심으로 최근 연구와 향후 연구 방향에 대해 논의하였습니다.&lt;/p&gt;
&lt;p&gt;방문 중 고종환 교수님은 세미나를 통해 IRIS Lab의 연구 방향과 관련 연구를 소개하였습니다. 이어진 공동 미팅에서는 각 연구진이 진행 중인 연구를 공유하고, AI 모델 경량화와 효율적인 인공지능 시스템 설계를 위한 연구 아이디어와 향후 발전 방향에 대해 의견을 나누었습니다.&lt;/p&gt;
&lt;p&gt;또한 연구진은 현재 진행 중인 공동 연구를 점검하고, model compression과 efficient AI systems 분야에서 향후 협력을 확대할 수 있는 연구 주제와 방향에 대해 논의하였습니다.&lt;/p&gt;
&lt;p&gt;이번 방문을 통해 양 연구진의 연구 현황과 관심 분야를 공유하고, efficient AI, model compression 및 HW–SW co-design 분야에서 지속적인 연구 교류와 협력 가능성을 모색하는 뜻깊은 기회를 가졌습니다.&lt;/p&gt;
&lt;p&gt;Under the supervision of Prof. Jong Hwan Ko, Ph.D. student Chanwoong Park and undergraduate researcher Seokho Han from the IRIS Lab visited the research lab led by Prof. Huanrui Yang at the University of Arizona on March 9, 2026, for a seminar and research exchange.&lt;/p&gt;
&lt;p&gt;The visit was organized to expand research exchange in efficient AI and AI model compression. In particular, discussions focused on shared research interests including model quantization, efficient AI inference, hardware-aware optimization, and HW–SW co-design.&lt;/p&gt;
&lt;p&gt;During the visit, Prof. Jong Hwan Ko gave a seminar introducing the research directions and related work of the IRIS Lab. This was followed by a joint meeting in which researchers from both labs shared their ongoing research and exchanged ideas and future directions on AI model compression and efficient AI system design.&lt;/p&gt;
&lt;p&gt;Furthermore, the researchers reviewed ongoing joint research and discussed potential research topics and directions to broaden future collaboration in model compression and efficient AI systems.&lt;/p&gt;
&lt;p&gt;Through this visit, the two labs shared their ongoing research and common interests and explored opportunities for sustained research exchange and collaboration in efficient AI, model compression, and HW–SW co-design.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C94] REFLEX: Rewrite-Free Row-Aligned Sparse Attention for Efficient LLM Execution on PIM</title>
      <link>https://iris-lab.skku.edu/publication/c94_dac_2026/</link>
      <pubDate>Thu, 26 Feb 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c94_dac_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Communications of the ACM Features IRIS Lab&#39;s Machine Unlearning Research</title>
      <link>https://iris-lab.skku.edu/post/interview_cacm/</link>
      <pubDate>Fri, 06 Feb 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/interview_cacm/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님 연구가 Communications of the ACM (CACM)에 소개되었습니다 ( &lt;a href=&#34;https://cacm.acm.org/news/teaching-ai-to-forget/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://cacm.acm.org/news/teaching-ai-to-forget/&lt;/a&gt; ). CACM은 ACM(Association for Computing Machinery)의 공식 매거진으로, 컴퓨터 과학 분야에서 가장 권위 있는 매체 중 하나입니다.&lt;/p&gt;
&lt;p&gt;프리랜스 과학기술 저널리스트 Sandrine Ceurstemont가 작성한 &amp;ldquo;Teaching AI to Forget&amp;rdquo;(AI에게 잊는 법을 가르치다) 기사는 머신 언러닝(Machine Unlearning) 기술을 다룹니다. 현재 AI 시스템은 단 몇 초의 음성 샘플만으로 목소리를 현실적으로 복제할 수 있습니다. 실제로 Oprah Winfrey의 목소리가 온라인 광고에 딥페이크로 사용되었고, 미국 고위 공직자의 목소리로 위조된 AI 음성 메시지가 개인 계정 접근에 악용되는 사례가 발생했습니다.&lt;/p&gt;
&lt;p&gt;이러한 문제를 해결하기 위해 AI 모델이 특정 음성을 &amp;ldquo;잊게&amp;rdquo; 만드는 것이 필요하지만, 전체 모델을 처음부터 재학습하는 것은 비용이 많이 듭니다. 머신 언러닝은 특정 데이터의 학습 영향을 효율적으로 제거하는 접근법으로, EU의 &amp;lsquo;잊힐 권리&amp;rsquo; 규정과 같은 데이터 프라이버시 법규에 대응할 수 있는 실용적인 해결책입니다.&lt;/p&gt;
&lt;p&gt;기사는 하버드대 Martin Pawelczyk 박사의 In-Context Unlearning 기법과 함께, 고종환 교수님의 음성 언러닝 연구를 주요 사례로 심층 소개합니다. 교수님 연구팀은 동의 없이 자신의 목소리가 AI로 복제되는 것을 원치 않는 사람들을 위한 솔루션을 개발했습니다. 기존의 가드레일 필터 방식과 달리, 이 방법은 특정 화자의 정체성을 영구적으로 숨기도록 AI 모델을 학습시켜, 해당 화자의 목소리를 생성하려 할 때마다 새로운 무작위 음성을 생성하게 합니다.&lt;/p&gt;
&lt;p&gt;연구팀은 Meta의 Voicebox 시스템을 대상으로 실험하여, 최대 10개의 목소리를 동시에 보호할 수 있음을 입증했습니다. 교수님은 &amp;ldquo;언러닝이 너무 강하면 모델이 다른 화자의 목소리를 생성하는 능력을 잃고, 너무 약하면 특정 목소리의 복제를 막을 수 없기 때문에 언러닝 과정이 매우 까다롭다&amp;quot;고 설명했습니다. 현재 연구팀은 더 많은 음성 제거 요청을 처리하고 처리 속도를 높이는 연구를 진행 중이며, 이 기술을 이미지나 비디오와 같은 다른 미디어로 확장하는 작업도 진행하고 있습니다.&lt;/p&gt;
&lt;p&gt;저희 연구는 지난 2025년 7월 MIT Technology Review에 이어 CACM에도 소개되어, 국제적으로 주목받고 있음을 보여줍니다. AI 기술이 발전할수록 사용자의 권리를 보호할 수 있는 메커니즘이 필요하다는 점에서, 머신 언러닝은 앞으로 AI 연구의 핵심 주제가 될 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;IRIS Lab&amp;rsquo;s research by Prof. Jong Hwan Ko has been featured in Communications of the ACM (CACM) ( &lt;a href=&#34;https://cacm.acm.org/news/teaching-ai-to-forget/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://cacm.acm.org/news/teaching-ai-to-forget/&lt;/a&gt; ). CACM is the flagship magazine of the Association for Computing Machinery (ACM) and one of the most prestigious publications in computer science.&lt;/p&gt;
&lt;p&gt;The article &amp;ldquo;Teaching AI to Forget,&amp;rdquo; written by freelance science and technology journalist Sandrine Ceurstemont, explores machine unlearning technology. AI systems can now realistically recreate voices from just a few seconds of speech samples. Real-world cases include Oprah Winfrey&amp;rsquo;s voice being deepfaked for online advertisements and AI-generated voice messages impersonating senior U.S. government officials to gain access to personal accounts.&lt;/p&gt;
&lt;p&gt;To address such problems, AI models need to &amp;ldquo;forget&amp;rdquo; certain voices, but retraining entire models from scratch is expensive. Machine unlearning offers a practical solution by efficiently removing the impact of specific training data, addressing data privacy regulations like the EU&amp;rsquo;s &amp;ldquo;right to be forgotten.&amp;rdquo;&lt;/p&gt;
&lt;p&gt;The article features Prof. Ko&amp;rsquo;s voice unlearning research as a major case study, alongside Harvard&amp;rsquo;s Dr. Martin Pawelczyk&amp;rsquo;s In-Context Unlearning approach. Prof. Ko&amp;rsquo;s team developed a solution for people who don&amp;rsquo;t want their voices cloned without consent. Unlike conventional guardrail filters, this method trains the AI model to permanently hide the identity of specific speakers by generating a new random voice each time it encounters a speaker who wants to be forgotten.&lt;/p&gt;
&lt;p&gt;The team tested their approach on Meta&amp;rsquo;s Voicebox system, demonstrating the ability to protect up to 10 voices simultaneously. Prof. Ko explains, &amp;ldquo;The unlearning process is quite tricky. If the unlearning is too strong, then the model can lose the ability to generate the remaining speakers&amp;rsquo; voices, and if it is too weak, then we cannot make the model refrain from reproducing specific voices.&amp;rdquo; The team is now working on handling more voice removal requests, speeding up performance, and adapting the technique for other media types such as images and videos.&lt;/p&gt;
&lt;p&gt;Following the MIT Technology Review feature in July 2025, this CACM article demonstrates the growing international recognition of our work. As AI technology advances, mechanisms for protecting user rights become essential, positioning machine unlearning as a key research area for the future of AI.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Workshop at KCS 2026</title>
      <link>https://iris-lab.skku.edu/post/workshop_kcs2026/</link>
      <pubDate>Fri, 30 Jan 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/workshop_kcs2026/</guid>
      <description>&lt;p&gt;IRIS 연구실은 2026년 1월 27일(화)부터 30일(금)까지 강원도 하이원리조트 그랜드 호텔 컨벤션타워에서 개최된 **제 33회 한국반도체학술대회(KCS 2026)**에 참가하며 연구실 워크샵을 진행하였습니다.&lt;/p&gt;
&lt;p&gt;KCS 2026은 &amp;ldquo;A Paradigm Shift in Semiconductors for AI Era&amp;quot;를 주제로, 서울대학교·한국반도체산업협회(KSIA) 등의 공동 주관 하에 열린 국내 대표 반도체 학술 행사입니다. AI 시대의 반도체 기술 패러다임 전환을 중심으로 활발한 연구 발표와 논의가 이루어졌으며, IRIS 연구실 구성원들도 최신 연구 동향을 폭넓게 접하는 기회를 가졌습니다.&lt;/p&gt;
&lt;p&gt;워크샵 기간 중에는 스키장·평창송어축제·동해 배낚시체험 등 강원도의 다양한 겨울 활동을 함께 즐기며 구성원들 간 교류와 친목을 다졌습니다. 한편, 고종환 교수님과 일부 구성원들은 ICML 논문 제출 마감을 앞두고 자유일정 시간에도 함께 논문 작업에 매진하여, 연구에 대한 열정과 팀워크를 다시 한번 확인할 수 있었습니다.&lt;/p&gt;
&lt;p&gt;마지막으로, 이번 워크샵을 아낌없이 지원해 주신 고종환 교수님과 기획·진행 전반을 맡아 수고해 주신 황찬욱 연구원을 비롯한 모든 구성원들께 진심으로 감사드립니다. 여러분의 노력 덕분에 학문적으로도, 인간적으로도 뜻깊은 시간이 될 수 있었습니다.&lt;/p&gt;
&lt;p&gt;IRIS Lab held its workshop in conjunction with the &lt;strong&gt;33rd Korean Conference on Semiconductors (KCS 2026)&lt;/strong&gt;, from January 27 to 30, 2026, at Highone Resort Grand Hotel Convention Tower in Gangwon Province.&lt;/p&gt;
&lt;p&gt;KCS 2026 was organized under the theme &amp;ldquo;A Paradigm Shift in Semiconductors for AI Era,&amp;rdquo; co-hosted by Seoul National University and the Korea Semiconductor Industry Association (KSIA), among others. Lab members had the opportunity to engage with the latest research trends in AI-era semiconductor technologies through presentations and discussions with leading researchers.&lt;/p&gt;
&lt;p&gt;During the workshop, members also enjoyed a range of winter activities in Gangwon Province—including skiing, the Pyeongchang Trout Festival, and sea fishing—strengthening bonds and creating lasting memories together. At the same time, Prof. Jong Hwan Ko and several members devoted part of the free schedule to collaborating on a paper for the ICML submission deadline, once again showcasing the lab&amp;rsquo;s dedication to research.&lt;/p&gt;
&lt;p&gt;We would like to extend our heartfelt gratitude to Prof. Jong Hwan Ko for his generous support throughout the workshop, and to researcher Chanwook Hwang and all the members who worked hard to plan and carry out the event. It was a truly meaningful experience, both academically and personally, thanks to everyone&amp;rsquo;s efforts.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Alumni Cheng Wencan (정문찬) Appointed as Professor at Sun Yat-sen University</title>
      <link>https://iris-lab.skku.edu/post/faculty_moonchan_jung/</link>
      <pubDate>Sun, 18 Jan 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/faculty_moonchan_jung/</guid>
      <description>&lt;p&gt;IRIS 연구실 박사과정 졸업생인 정문찬(Cheng Wencan) 연구원이 2026년 4월부터 중국 Sun Yat-sen 대학교(Sun Yat-sen University) 인공지능학과 교수로 임용됩니다.&lt;/p&gt;
&lt;p&gt;정문찬 박사는 성균관대학교에서 박사학위를 취득한 후, 싱가포르국립대학교(National University of Singapore, NUS)에서 박사후연구원으로 활동하며 효율적인 하드웨어 및 3D 구조 데이터 처리 연구를 수행해왔습니다.&lt;/p&gt;
&lt;p&gt;IRIS 연구실에서 정문찬 박사는 3D 손 자세 추정, 포인트 클라우드 기반 장면 흐름 추정, FPGA 기반 실시간 시스템 등 다양한 연구 성과를 이루었으며, IEEE IoT Journal을 비롯한 저명한 학술지와 CVPR, ICCV 등 최고 수준의 국제 학회에 다수의 논문을 발표했습니다.&lt;/p&gt;
&lt;p&gt;주요 연구 업적으로는 Diffusion 모델을 활용한 3D 손 자세 추정(HandDiff), 적응형 그래프 트랜스포머 기반 노이즈 제거 기법(HandDAGT), 간접 ToF 센서를 위한 FPGA 기반 에너지 효율적 실시간 시스템 등이 있으며, 이러한 연구들은 효율적인 3D 비전 시스템 구현에 중요한 기여를 했습니다.&lt;/p&gt;
&lt;p&gt;IRIS 연구실은 정문찬 박사의 교수 임용을 진심으로 축하하며, 앞으로 Sun Yat-sen 대학교에서 인공지능 분야의 우수한 연구자와 교육자로서 더욱 발전하기를 기원합니다.&lt;/p&gt;
&lt;p&gt;IRIS Lab is delighted to announce that Dr. Cheng Wencan (정문찬), a distinguished Ph.D. graduate from our laboratory, has been appointed as a Professor in the Department of Artificial Intelligence at Sun Yat-sen University, effective April 2026.&lt;/p&gt;
&lt;p&gt;After completing his Ph.D. at Sungkyunkwan University, Dr. Cheng served as a Postdoctoral Researcher at the National University of Singapore (NUS), where he conducted research on efficient hardware and 3D structural data processing.&lt;/p&gt;
&lt;p&gt;During his time at IRIS Lab, Dr. Cheng made significant contributions to various research areas, including 3D hand pose estimation, point cloud-based scene flow estimation, and FPGA-based real-time systems. His work has been published in prestigious journals such as the IEEE IoT Journal and presented at top-tier international conferences including CVPR and ICCV.&lt;/p&gt;
&lt;p&gt;His notable research achievements include HandDiff (3D hand pose estimation using diffusion models), HandDAGT (denoising with adaptive graph transformers), and an FPGA-based energy-efficient real-time system for indirect ToF sensors. These contributions have significantly advanced the field of efficient 3D vision systems.&lt;/p&gt;
&lt;p&gt;IRIS Lab extends its warmest congratulations to Dr. Cheng Wencan on his faculty appointment and wishes him continued success as a leading researcher and educator in artificial intelligence at Sun Yat-sen University.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[J36] RUnQuant: High-Resolution Weight Quantization via Unanchored Weight Decomposition in Column-wise Granularity for CIM Accelerators</title>
      <link>https://iris-lab.skku.edu/publication/j36_jsa_2026/</link>
      <pubDate>Sat, 17 Jan 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j36_jsa_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C93] SODA: A Unified Framework for Join Estimation of Speaker Orientation and Direction of Arrival</title>
      <link>https://iris-lab.skku.edu/publication/c93_icassp_2026/</link>
      <pubDate>Fri, 16 Jan 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c93_icassp_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Prof. Jong Hwan Ko Honored with 2025 SKKU Rising-Fellowship</title>
      <link>https://iris-lab.skku.edu/post/award_skku_rising_fellowship/</link>
      <pubDate>Thu, 15 Jan 2026 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/award_skku_rising_fellowship/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님이 성균관대학교가 올해 신설한 &amp;lsquo;2025 SKKU Rising-Fellowship&amp;rsquo; 교수로 선정되었습니다.&lt;/p&gt;
&lt;p&gt;SKKU Rising-Fellowship은 본교 신진·중견 전임교원 중 해당 학문 분야에서 국내 최고 수준 또는 세계적 표준에 이미 안착하였거나, 연구업적이 탁월하여 향후 세계적 수준의 연구자로 발전 가능성이 높은 최우수 교원에게 수여되는 명예이자 특별연구 지원 제도입니다.&lt;/p&gt;
&lt;p&gt;2025년도 수상자는 연구 성과의 학문적·질적 우수성과 글로벌 연구 영향력을 종합적으로 고려하여 선정위원회의 엄정한 심사를 거쳐 선정되었으며, 고종환 교수님을 포함한 총 17명의 교수가 이 영예를 안았습니다.&lt;/p&gt;
&lt;p&gt;시상식은 2025년 12월 23일 화요일에 진행되었으며, 유지범 총장 및 학교법인 관계자, 교내 보직교수가 참석하여 수상자들을 축하하고 격려의 메시지를 전했습니다.&lt;/p&gt;
&lt;p&gt;고종환 교수님은 자원 효율적 AI 시스템, 하드웨어-소프트웨어 공동 설계, 인-메모리 컴퓨팅 등 다양한 분야에서 세계적 수준의 연구 성과를 창출해 왔으며, 이번 수상은 그간의 탁월한 연구 업적과 글로벌 영향력을 인정받은 결과입니다.&lt;/p&gt;
&lt;p&gt;IRIS 연구실 구성원 모두는 교수님의 수상을 진심으로 축하드리며, 앞으로도 더욱 훌륭한 연구 성과를 이어가실 것을 기대합니다.&lt;/p&gt;
&lt;p&gt;자세한 내용은 성균관대학교 공식 보도자료에서 확인하실 수 있습니다: &lt;a href=&#34;https://www.skku.edu/skku/campus/skk_comm/news.do?mode=view&amp;amp;articleNo=133604&amp;amp;article.offset=0&amp;amp;articleLimit=10&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;성균관대학교 뉴스&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Prof. Jong Hwan Ko from IRIS Lab has been selected as a recipient of the newly established &amp;lsquo;2025 SKKU Rising-Fellowship&amp;rsquo; by Sungkyunkwan University.&lt;/p&gt;
&lt;p&gt;The SKKU Rising-Fellowship is a prestigious honor and special research support program awarded to outstanding junior and mid-career faculty members who have already established themselves at the highest domestic level or world-class standards in their academic fields, or whose research achievements are exceptional and show high potential for development as world-class researchers.&lt;/p&gt;
&lt;p&gt;The 2025 recipients were selected through rigorous evaluation by the selection committee, considering the academic and qualitative excellence of research achievements and global research impact. A total of 17 professors, including Prof. Jong Hwan Ko, were honored with this distinction.&lt;/p&gt;
&lt;p&gt;The award ceremony was held on Tuesday, December 23, 2025, with President Yoo Ji-bum, school foundation officials, and university administrators in attendance to congratulate and encourage the recipients.&lt;/p&gt;
&lt;p&gt;Prof. Jong Hwan Ko has produced world-class research outcomes in various fields including resource-efficient AI systems, hardware-software co-design, and in-memory computing. This award recognizes his outstanding research achievements and global impact.&lt;/p&gt;
&lt;p&gt;All members of IRIS Lab sincerely congratulate Professor Ko on this honor and look forward to his continued excellence in research.&lt;/p&gt;
&lt;p&gt;For more details, please visit the official Sungkyunkwan University news release: &lt;a href=&#34;https://www.skku.edu/skku/campus/skk_comm/news.do?mode=view&amp;amp;articleNo=133604&amp;amp;article.offset=0&amp;amp;articleLimit=10&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;SKKU News&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>NeurIPS 2025 and West Coast Lab Visits</title>
      <link>https://iris-lab.skku.edu/post/us_west/</link>
      <pubDate>Fri, 26 Dec 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/us_west/</guid>
      <description>&lt;p&gt;우리 연구실은 2025년 12월 NeurIPS 2025 학회 참여와 더불어 미국 서부의 주요 연구 기관인 UCSD, Samsung Research, UCLA를 방문해 초청 강연과 연구 교류를 진행했습니다.&lt;/p&gt;
&lt;p&gt;먼저 미국 샌디에고에서 열린 NeurIPS (Neural Information Processing Systems) 2025 학회 메인 세션에 승인된 3편의 논문(“Optimization Minimal 3D Gaussian Splatting”, “Efficient Multi-bit Quantization Network Training via Weight Bias Correction and Bit-wise Coreset Sampling”, “Perturb a Model, Not an Image: Towards Robust Privacy Protection via Anti-Personalized Diffusion Models”)과 워크샵 세션에 승인된 2편의 논문(“No Encore: Unlearning as Opt-Out in Music Generation”, “Do Not Mimic My Voice : Speaker Identity Unlearning for Zero-Shot Text-to-Speech”)의 포스터 및 워크샵 발표를 진행했습니다.&lt;/p&gt;
&lt;p&gt;이후 University of California San Diego (UCSD)의 Tajana Rosing, 강민구 교수님 연구실에 방문하여 에너지 효율적 AI 시스템을 위해 인메모리 컴퓨팅 기반 가속기와 엣지 환경 최적화 연구를 공통 축으로 협력 가능성을 논의했으며, 삼성 Labs 샌디에고에 방문하여 On-device AI 주제의 초청 강연을 진행했습니다.&lt;/p&gt;
&lt;p&gt;마지막으로 University of California Los Angeles (UCLA)의 Cho-Jui Hsieh, Baharan Mirzasoleiman 교수 연구실에 방문하여 대규모 AI를 위한 최적화·데이터마이닝 기반 알고리즘을 중심으로 협력 가능성을 논의하며 뜻깊은 시간을 보냈습니다.&lt;/p&gt;
&lt;p&gt;이번 방문을 따뜻하게 맞아 주신 Tajana Rosing 교수님, 강민구 교수님, Burhan A. Mudassar 박사님, Cho-Jui Hsieh 교수님, Baharan Mirzasoleiman 교수님께 깊이 감사드립니다.&lt;/p&gt;
&lt;p&gt;Our lab participated in NeurIPS 2025 in December 2025, and additionally visited major research institutions on the U.S. West Coast—UCSD, Samsung Research, and UCLA—to conduct invited talks and research exchanges.&lt;/p&gt;
&lt;p&gt;First, at NeurIPS (Neural Information Processing Systems) 2025 held in San Diego, USA, we delivered poster and workshop presentations for three papers accepted to the main conference sessions (“Optimization Minimal 3D Gaussian Splatting,” “Efficient Multi-bit Quantization Network Training via Weight Bias Correction and Bit-wise Coreset Sampling,” and “Perturb a Model, Not an Image: Towards Robust Privacy Protection via Anti-Personalized Diffusion Models”) and two papers accepted to the workshop sessions (“No Encore: Unlearning as Opt-Out in Music Generation” and “Do Not Mimic My Voice: Speaker Identity Unlearning for Zero-Shot Text-to-Speech”).&lt;/p&gt;
&lt;p&gt;Afterward, we visited the labs of Professors Tajana Rosing and Mingu Kang at the University of California San Diego (UCSD), where we discussed potential collaboration based on our shared focus on in-memory-computing-based accelerators and edge-environment optimization research for energy-efficient AI systems, and we visited Samsung Labs at San Diego to deliver an invited talk on on-device AI.&lt;/p&gt;
&lt;p&gt;Finally, we visited the labs of Professors Cho-Jui Hsieh and Baharan Mirzasoleiman at the University of California, Los Angeles (UCLA), where we spent a meaningful time discussing potential collaboration centered on optimization- and data-mining-based algorithms for large-scale AI.&lt;/p&gt;
&lt;p&gt;We sincerely thank Professor Tajana Rosing, Professor Mingu Kang, Dr. Burhan A. Mudassar, Professor Cho-Jui Hsieh, and Professor Baharan Mirzasoleiman for warmly welcoming us during this visit.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>ICCAD 2025 and Europe Lab Visits</title>
      <link>https://iris-lab.skku.edu/post/eu/</link>
      <pubDate>Tue, 23 Dec 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/eu/</guid>
      <description>&lt;p&gt;IRIS 연구실의 전강은 박사후연구원 인솔 하에 이존이, 강도영 학생은 2025년 10월 26일부터 2025년 11월 1일까지 ICCAD 2025 학회 참여와 더불어 유럽 주요 연구 기관인 Politecnico di Torino와 ETH Zurich를 방문해 연구 교류를 진행했습니다.&lt;/p&gt;
&lt;p&gt;먼저 독일 뮌헨에서 열린 ICCAD (International Conference on Computer-Aided Design) 2025 학회에서 전강은 박사의 논문 (제목: Row-Column Hybrid Grouping for Fault-Resilient Multi-Bit Weight Representation on IMC Arrays) 발표와 함께 이존이 학생의 Student Research Competition (주제: Optimizing Weight Mapping for Energy-Efficient In-Memory CNN Inference)에서 포스터/오랄 발표를 진행했습니다.&lt;/p&gt;
&lt;p&gt;이후 이탈리아 토리노 공대(Politecnico di Torino) Daniele Paliari, Alessio Burrello 교수 연구실에 방문하여 양 연구실의 연구 협력 방안을 공유하고, Neural Architecture Search, 모델 추론 최적화 등의 분야에서 협력 가능성을 논의하였습니다.&lt;/p&gt;
&lt;p&gt;또한 스위스 취리히 연방 공대(ETH Zurich) Ana Klimovic 교수 연구실에 방문하여 Operating System (OS)단에서 양자화 적용 등 효율적인 LLM Serving과 관련하여 협력 가능성을 논의하였습니다. 마지막으로 IBM Analog IMC Group에 Julian Buchel 연구원과 만나 파운데이션 모델의 Analog IMC 가속화 분야의 최신 연구 동향에 대한 견해를 주고 받으며 뜻깊은 시간을 보냈습니다.&lt;/p&gt;
&lt;p&gt;Under the supervision of Dr. Kang Eun Jeon (Postdoctoral Researcher), Johnny Rhe and Do Yeong Kang from the IRIS Lab participated in ICCAD 2025 and visited major European research institutions—Politecnico di Torino and ETH Zurich—for research exchange from October 26 to November 1, 2025.&lt;/p&gt;
&lt;p&gt;At ICCAD 2025 in Munich, Germany, Dr. Kang Eun Jeon presented his paper titled “Row-Column Hybrid Grouping for Fault-Resilient Multi-Bit Weight Representation on IMC Arrays,” and Johnny Rhe delivered both a poster and an oral presentation in the Student Research Competition on “Optimizing Weight Mapping for Energy-Efficient In-Memory CNN Inference.”&lt;/p&gt;
&lt;p&gt;After the conference, we visited Prof. Daniele Paliari and Prof. Alessio Burrello’s lab at Politecnico di Torino to share potential collaboration plans and discuss joint research opportunities in areas such as Neural Architecture Search and inference optimization.&lt;/p&gt;
&lt;p&gt;We also visited Prof. Ana Klimovic’s lab at ETH Zurich to explore collaboration on efficient LLM serving, including OS-level quantization techniques. Finally, we met with researcher Julian Buchel at IBM’s Analog IMC Group to exchange insights on the latest trends in analog IMC acceleration for foundation models.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C92] Harnessing Linguistic Dissimilarity for Language Generalization on Unseen Low-Resource Varieties</title>
      <link>https://iris-lab.skku.edu/publication/c92_aaai_workshop_2026/</link>
      <pubDate>Tue, 16 Dec 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c92_aaai_workshop_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C91] No Encore: Unlearning as Opt-Out in Music Generation</title>
      <link>https://iris-lab.skku.edu/publication/c91_neurips_workshop_2025/</link>
      <pubDate>Mon, 15 Dec 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c91_neurips_workshop_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J35] RUSH: Recursive and Scalable 3D Coarse To Fine Path Planning</title>
      <link>https://iris-lab.skku.edu/publication/j35_ieee-ra-l_2026/</link>
      <pubDate>Sun, 14 Dec 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j35_ieee-ra-l_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Excellence Award at the SKKU Graduate Student Paper Competition (Researcher: Hyeonsu Bang)</title>
      <link>https://iris-lab.skku.edu/post/award_skku_graduate/</link>
      <pubDate>Wed, 10 Dec 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/award_skku_graduate/</guid>
      <description>&lt;p&gt;IRIS 연구실 박사과정 방현수 연구원이 2025년 SKKU 대학원생 논문대상에서 우수상을 수상했습니다.&lt;/p&gt;
&lt;p&gt;수상 논문의 제목은 “Overhead-Free Weight Remapping for Efficient Fault Tolerance in MLC ReRAM Arrays” 로, MLC ReRAM 기반 AI 가속기의 신뢰성과 효율성을 동시에 확보할 수 있는 새로운 결함 내성 기법을 제안한 연구입니다.&lt;/p&gt;
&lt;p&gt;이번 수상은 메모리 기반 컴퓨팅 구조에서의 신뢰성 향상에 대한 연구의 우수성과 잠재력을 인정받은 결과로, 향후 관련 분야 연구 발전에도 의미 있는 기여를 할 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;현재 해당 연구는 저널에 under review 상태이며, 방현수 연구원은 앞으로도 메모리 기반 AI 컴퓨팅의 실용화와 고도화를 위한 연구를 지속할 계획입니다.&lt;/p&gt;
&lt;p&gt;Our Ph.D. student Hyeonsu Bang has received the Excellence Award at the 2025 SKKU Graduate Student Paper Competition.&lt;/p&gt;
&lt;p&gt;His award-winning paper, titled “Overhead-Free Weight Remapping for Efficient Fault Tolerance in MLC ReRAM Arrays,” proposes a novel fault-tolerance scheme that enhances both reliability and efficiency in ReRAM–based AI accelerator.&lt;/p&gt;
&lt;p&gt;This award recognizes the significance and potential impact of his research in advancing high-reliability, memory-centric computing architectures. It also highlights the contribution his work is expected to make to future developments in the field.&lt;/p&gt;
&lt;p&gt;The paper is currently under review in a journal, and he plans to continue pursuing research aimed at further advancing and practicalizing memory-based AI computing.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C90] Single-step Diffusion for Image Compression at Ultra-Low Bitrates</title>
      <link>https://iris-lab.skku.edu/publication/c90_wacv_2026/</link>
      <pubDate>Tue, 25 Nov 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c90_wacv_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Carnegie Mellon University (CMU) Collaboration (Visiting Researcher: Jinju Kim)</title>
      <link>https://iris-lab.skku.edu/post/visiting_researcher_jinju_kim/</link>
      <pubDate>Thu, 02 Oct 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/visiting_researcher_jinju_kim/</guid>
      <description>&lt;p&gt;IRIS 연구실 석사과정 김진주 연구원 (&lt;a href=&#34;https://mokcho.github.io/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;mokcho.github.io&lt;/a&gt;)은 2024년 8월부터 카네기멜론대학교 (Carnegie Mellon University, 이하 CMU) 에 방문 연구원으로 활동하며, 음성과 오디오 차원에서 머신 언러닝에 대한 연구를 세계적 전문가들과 함께 긴밀히 이어가고 있습니다.&lt;/p&gt;
&lt;p&gt;김진주 연구원의 CMU 여정은 서강대학교와 한국정보통신기획평가원 (IITP)의 지원을 받아 6개월간의 연구 프로그램으로 시작되었으며, 교수진과의 교류를 통해 향후 지속적 협업 기반을 다졌습니다.&lt;/p&gt;
&lt;p&gt;공식 프로그램이 종료된 이후에도, 김진주 연구원은 원격으로 David R. Mortensen 교수님과 연구를 지속하며 ACL 언어학회들에 논문을 제출하고, Rita Singh 교수님과 함께 2025 NeurIPS AI4Music 워크숍에서 음악 생성 모델에 대한 머신 언러닝 논문 (&lt;em&gt;“No Encore: Unlearning as Opt-Out in Music Generation”&lt;/em&gt;) 을 1저자로 발표가게 되었습니다. 이 성과를 바탕으로 Bhiksha Raj 교수님의 연구실 MLSP Group에서 음성 및 오디오 시스템에 언러닝 기법을 적용한 보안 연구를 CMU 현지에서 지속하자는 초청을 받았습니다.&lt;/p&gt;
&lt;p&gt;2025년 9월부터 김진주 연구원은 Bhiksha Raj 교수님, Rita Singh 교수님과 함께 &amp;ldquo;특정 데이터로 학습한 정보를 잊는 효율적인 알고리즘 설계&amp;quot;에 집중하고 있습니다. 이 주제는 음성 프라이버시, 윤리적 AI 생성 시스템, 효율적 머신러닝 등 여러 분야와 깊이 연결되어 있습니다.&lt;/p&gt;
&lt;p&gt;김진주 연구원의 CMU 방문은 한국 대학과 세계적 AI 연구 거점 간의 연결을 강화하며, 국제 공동 연구의 시너지를 만드는 기회가 될 것입니다. 앞으로 머신 언러닝, 음성/프라이버시 시스템 분야에서 연구의 결실을 기대합니다.&lt;/p&gt;
&lt;p&gt;Since August 2024, our Master&amp;rsquo;s student Jinju Kim has been engaging as a visiting research fellow in the School of Computer Science at Carnegie Mellon University. Since then, Jinju Kim has evolved her research on machine unlearning into close collaboration with leading experts in language, audio, and privacy.&lt;/p&gt;
&lt;p&gt;Her journey with CMU began through a six-month funded program sponsored by Sogang University and the Institute for Information &amp;amp; Communications Technology Planning &amp;amp; Evaluation (IITP). During that period, Jinju Kim immersed herself in four courseworks, engaged with faculty, and laid the groundwork for long-term collaboration.&lt;/p&gt;
&lt;p&gt;After the official program ended, Jinju Kim continued collaborating remotely with CMU professors, submitting work with Prof. David R. Mortensen to ACL venues and leading as first author on a paper with Prof. Rita Singh to be presented at the NeurIPS 2025 AI4Music workshop titled “No Encore: Unlearning as Opt-Out in Music Generation.” Recognizing the promise of this work, Prof. Bhiksha Raj’s lab, the MLSP Group, invited Jinju Kim to continue her research on-site, focusing on efficient machine unlearning methods for audio and speech systems.&lt;/p&gt;
&lt;p&gt;From September 2025, Jinju Kim’s ongoing work centers on designing unlearning algorithms that can “forget targeted data distributions efficiently.” This is especially relevant to voice privacy, trustworthy generative AI systems, and efficient machine learning.&lt;/p&gt;
&lt;p&gt;Jinju Kim’s presence at CMU strengthens ties between Korean universities and a leading AI research hub, catalyzing new synergies across international and interdisciplinary research. We look forward to the full impact of her evolving work.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Duke University Collaboration (Visiting Researcher: Juhong Park)</title>
      <link>https://iris-lab.skku.edu/post/visiting_researcher_joohong_park/</link>
      <pubDate>Thu, 02 Oct 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/visiting_researcher_joohong_park/</guid>
      <description>&lt;p&gt;IRIS 연구실 석박통합과정 박주홍 연구원은 현재 미국 Duke University의 CEI Group (Prof. Yiran Chen)에서 Visiting Scholar로 연구를 수행하고 있습니다.&lt;/p&gt;
&lt;p&gt;박주홍 연구원의 연구는 이종 인메모리 컴퓨팅 환경에서 거대 언어 모델(LLM)을 효율적으로 실행하기 위한 하드웨어–소프트웨어 최적화에 초점을 두고 있습니다. 특히 DRAM-PIM(Processing-in-Memory)과 CIM(Compute-in-Memory)과 같은 신흥 메모리 중심 아키텍처를 대상으로, 압축 기법의 적용, 이종 환경에서의 워크로드 스케줄링, 컴파일러 기법을 포함한 Near-Data Processing(NDP) 기반 공동 최적화 전략을 탐구하고 있습니다.&lt;/p&gt;
&lt;p&gt;이를 통해 연산·메모리·전력 제약이 큰 환경에서도 실용적으로 적용 가능한 자원 효율적 LLM 서빙 경로를 제시하고 있으며, 앞으로도 CEI Group과의 협력을 통해 LLM 시스템 최적화와 메모리 지향 아키텍처의 융합을 지속적으로 확장해 나갈 예정입니다.&lt;/p&gt;
&lt;p&gt;Juhong Park, a Combined M.S./Ph.D. student in our lab, is currently serving as a Visiting Scholar at Duke University’s CEI Group (Prof. Yiran Chen).&lt;/p&gt;
&lt;p&gt;Her research focuses on hardware–software co-optimization for efficient large language model (LLM) execution in heterogeneous in-memory computing environments. In particular, she explores joint optimization strategies in Near-Data Processing (NDP) systems, targeting emerging memory-centric architectures such as DRAM-PIM (Processing-in-Memory) and CIM (Compute-in-Memory). Her work includes the efficient application of compression techniques, workload scheduling across heterogeneous systems, and compiler-level optimization methods.&lt;/p&gt;
&lt;p&gt;Through this research, she aims to propose resource-efficient LLM serving pathways that remain practical under compute, memory, and power constraints. We look forward to her continued efforts to advance the convergence of LLM system optimization and memory-centric architectures in collaboration with the CEI Group.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>East Coast Lab Visits and Talks: Resource-Efficient AI via SW-HW Co-design</title>
      <link>https://iris-lab.skku.edu/post/us/</link>
      <pubDate>Mon, 29 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/us/</guid>
      <description>&lt;p&gt;우리 연구실은 2025년 8월 미국 동부의 주요 기관인 Duke University, Cornell Tech, University of Illinois Chicago, University of North Carolina를 방문해 초청 강연과 연구 교류를 진행했습니다.&lt;/p&gt;
&lt;p&gt;강연 주제는 ‘SW–HW 공동 설계를 통한 자원 효율적 인공지능 구현(Enabling Resource-Efficient AI via SW–HW Co-design)’이며, 혼합·다중 정밀도 양자화를 비롯한 압축 기법이 인-메모리 컴퓨팅 등 신흥 하드웨어 패러다임과 어떻게 공동 최적화될 수 있는지를 다뤘습니다. 이를 통해 연산·메모리·전력 제약이 있는 환경에서도 실용적으로 배치 가능한 AI 경로를 제안했습니다.&lt;/p&gt;
&lt;p&gt;강연 전체 영상은 Duke Athena Seminar Series에서 확인하실 수 있습니다.
전체 녹화본: &lt;a href=&#34;https://www.youtube.com/watch?v=_Krud-njeog&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.youtube.com/watch?v=_Krud-njeog&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;이번 방문을 따뜻하게 맞아 주신 Yiran Chen 교수님, 서재선 교수님, Amit Trivedi 교수님, Roni Sengupta 교수님께 깊이 감사드립니다. 또한 전 일정을 매끄럽게 조율해 주신 전강은 박사님과 준비·운영을 도와준 우리 연구실의 이존이, 이주찬, 연상흠 학생들에게 고마움을 전합니다.&lt;/p&gt;
&lt;p&gt;In August 2025, our lab visited major institutions on the U.S. East Coast—Duke University, Cornell Tech, the University of Illinois Chicago, and the University of North Carolina—to deliver invited talks and conduct research exchanges.&lt;/p&gt;
&lt;p&gt;The talks were on “Enabling Resource-Efficient AI via SW–HW Co-design,” addressing how compression techniques—including mixed- and multi-precision quantization—can be co-optimized with emerging hardware paradigms such as in-memory computing. Through this, we proposed AI deployment pathways that are practically feasible even under compute, memory, and power constraints.&lt;/p&gt;
&lt;p&gt;The full talk video is available from the Duke Athena Seminar Series.
Full recording: &lt;a href=&#34;https://www.youtube.com/watch?v=_Krud-njeog&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.youtube.com/watch?v=_Krud-njeog&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;We thank Prof. Yiran Chen, Prof. Jae-sun Seo, Prof. Amit Trivedi, and Prof. Roni Sengupta for their warm hospitality; Dr. Kang Eun JEON for seamlessly coordinating every detail; and our students Johnny Rhe, Joo Chan Lee, and Sang Heum Yeon for their tremendous support.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Undergraduate Researchers Lead ICCV 2025 Paper on Memory-Efficient Quantization</title>
      <link>https://iris-lab.skku.edu/post/iccv2025_undergraduated/</link>
      <pubDate>Mon, 15 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/iccv2025_undergraduated/</guid>
      <description>&lt;p&gt;IRIS 연구실의 학부연구생 한석호 학생(시스템경영공학과 3학년)과 윤서연 학생(전자전기공학부 졸업)이 1저자로 참여한 논문이 ICCV 2025(International Conference on Computer Vision)에 발표 승인되었습니다.&lt;/p&gt;
&lt;p&gt;본 연구는 학부생들이 아이디어 기획부터 실험 설계, 결과 분석, 논문 작성에 이르기까지 연구 전반을 주도하여 수행했습니다. 미국 University of Arizona의 Huanrui Yang 교수 연구팀과의 국제 공동 연구로 진행되었으며, IRIS 연구실의 글로벌 연구 네트워크를 통한 협력 성과입니다.&lt;/p&gt;
&lt;p&gt;논문 &amp;ldquo;MSQ: Memory-Efficient Bit Sparsification Quantization&amp;quot;은 모바일이나 엣지 디바이스처럼 자원이 제한된 환경에서 인공지능 모델을 효율적으로 학습시키기 위한 양자화 기법을 제안합니다. 기존의 비트 단위 분해 양자화는 정밀도 탐색이 유연하지만 학습 과정에서 메모리 사용량이 크게 증가하는 문제가 있었습니다.&lt;/p&gt;
&lt;p&gt;MSQ는 비트 분해 없이 중요도가 낮은 비트만을 선택적으로 제거하는 방식으로, 학습 중 메모리와 연산량을 동시에 절감합니다. 모델 민감도 정보를 반영하여 여러 개의 LSB를 한 번에 제거할 수 있도록 설계되어, 메모리 효율성과 학습 속도를 모두 개선했습니다.&lt;/p&gt;
&lt;p&gt;실험 결과 MSQ는 기존 방법 대비 메모리 효율성과 학습 속도에서 우수한 성능을 보였으며, 제한된 자원 환경에서도 효과적인 양자화 모델 학습이 가능함을 확인했습니다. IRIS 연구실은 앞으로도 학부생들의 주도적인 연구 참여를 지원하며, 글로벌 협력을 통해 온디바이스 AI 분야의 연구 성과를 지속적으로 창출해 나갈 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;자세한 내용은 정보통신대학 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=206415&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=206415&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab undergraduate researchers Seokho Han (3rd year, Systems Management Engineering) and Seoyeon Yoon (graduated, School of Electronic and Electrical Engineering) have had their first-author paper accepted at ICCV 2025 (International Conference on Computer Vision).&lt;/p&gt;
&lt;p&gt;The undergraduate students led the entire research process, from initial ideation through experimental design, result analysis, and paper writing. This research was conducted in collaboration with Prof. Huanrui Yang&amp;rsquo;s team at the University of Arizona, representing a collaborative achievement through IRIS Lab&amp;rsquo;s global research network.&lt;/p&gt;
&lt;p&gt;The paper &amp;ldquo;MSQ: Memory-Efficient Bit Sparsification Quantization&amp;rdquo; proposes a quantization technique for efficiently training AI models in resource-constrained environments such as mobile or edge devices. While existing bit-level splitting quantization offers flexible precision search, it significantly increases memory usage during training.&lt;/p&gt;
&lt;p&gt;MSQ selectively removes only low-importance bits without bit decomposition, simultaneously reducing both memory and computation during training. By reflecting model sensitivity information to remove multiple LSBs at once, it improves both memory efficiency and training speed.&lt;/p&gt;
&lt;p&gt;Experimental results showed that MSQ demonstrated superior performance in memory efficiency and training speed compared to existing methods, confirming effective quantized model training even in resource-constrained environments. IRIS Lab will continue to support undergraduate students&amp;rsquo; leading research participation and is expected to consistently produce research outcomes in on-device AI through global collaboration.&lt;/p&gt;
&lt;p&gt;For more details, please visit the official news from the College of Information &amp;amp; Communication Engineering: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=206415&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=206415&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Lab Photo Shoot</title>
      <link>https://iris-lab.skku.edu/post/photos/</link>
      <pubDate>Thu, 11 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/photos/</guid>
      <description>&lt;p&gt;전강은 박사님 주도하에 연구실 사진 촬영을 진행하여, 교수님과 연구실 구성원들의 새로운 프로필 및 단체 사진을 담았습니다. 우리 연구실을 기록하고 기념하는 뜻깊은 시간이었습니다. 주도와 원활한 조율을 해주신 전강은 박사님과 참여해 주신 모든 분들께 진심으로 감사드립니다.&lt;/p&gt;
&lt;p&gt;We conducted a lab photo shoot, led by Dr. Kang Eun Jeon, capturing new profile and group photos of the professor and all lab members. It was a meaningful time to document and celebrate our lab. We sincerely thank Dr. Kang Eun Jeon for the leadership and seamless coordination, and all participants for their support.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Prof. Yani Ioannou’s Seminar on ‘Training Structured Sparse Neural Networks’</title>
      <link>https://iris-lab.skku.edu/post/seminar_yani/</link>
      <pubDate>Wed, 10 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/seminar_yani/</guid>
      <description>&lt;p&gt;2025년 9월 10일, University of Calgary의 yani 교수님이 성균관대학교를 방문해 “Training Structured Sparse Neural Networks”라는 주제로 세미나를 진행해 주셨습니다. 세미나는 반도체관 400112호에서 열렸고, IRIS 연구실의 고종환 교수님이 호스트를 맡았습니다.&lt;/p&gt;
&lt;p&gt;발표에서는 동적 희소 학습(DST)의 개념과 한계를 정리한 뒤, 학습 단계에서부터 N:M 형태와 같은 하드웨어 친화적 구조적 희소성을 확보하는 Structured RigL(SRigL) 접근을 소개했습니다. 특히 뉴런 어블레이션을 명시적으로 통합해 매우 높은 희소화 수준에서도 일반화 성능을 유지하고, 실제 CPU/GPU 환경에서 추론 속도 향상을 확인한 점이 인상적이었습니다. 이번 세미나를 통해 구조적 희소 신경망 훈련의 배경과 설계상의 고려 요소를 한층 깊이 이해할 수 있었습니다.&lt;/p&gt;
&lt;p&gt;간단한 약력으로, yani 교수님은 University of Calgary Schulich School of Engineering의 전기·소프트웨어공학과에서 재직 중이며 Schulich Research Chair로서 Calgary Machine Learning Lab을 이끌고 있습니다. 이전에는 Vector Institute와 University of Guelph에서 박사후연구원을 지냈고, Google Brain Toronto에서 연구를 수행했습니다. 박사학위는 University of Cambridge에서 취득했으며, 효율적이고 신뢰할 수 있는 딥러닝, 특히 희소 신경망의 학습과 추론에 집중하고 있습니다.&lt;/p&gt;
&lt;p&gt;바쁜 일정에도 귀한 시간을 내어주신 Yani 교수님께 깊이 감사드립니다.&lt;/p&gt;
&lt;p&gt;On September 10, 2025, Prof. Yani Ioannou from the University of Calgary visited Sungkyunkwan University and delivered a seminar titled “Training Structured Sparse Neural Networks.” The talk was held in the Semiconductor Building, Room 400112, and hosted by Prof. Jong Hwan Ko from the IRIS Lab.&lt;/p&gt;
&lt;p&gt;The seminar first reviewed the ideas and practical limitations behind Dynamic Sparse Training (DST), then introduced Structured RigL (SRigL), which learns hardware-friendly N:M structured sparsity during training. A notable point was the explicit use of neuron ablation, enabling strong generalization even at extreme sparsity while demonstrating measurable inference speedups on CPUs and GPUs. The session offered a clear, in-depth view of how to design and train structured sparse neural networks that are both accurate and efficient.&lt;/p&gt;
&lt;p&gt;By way of a short bio, Prof. Ioannou is an Assistant Professor and a Schulich Research Chair in the Schulich School of Engineering at the University of Calgary, where he leads the Calgary Machine Learning Lab. He previously worked as a Postdoctoral Fellow at the Vector Institute and the University of Guelph and conducted research at Google Brain Toronto. He earned his Ph.D. from the University of Cambridge, and his research focuses on efficient and trustworthy deep learning, with an emphasis on training and inference for sparse neural networks.&lt;/p&gt;
&lt;p&gt;We sincerely thank Prof. Yani Ioannou for sharing his time and insights.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C89] In-Sensor Denoising Using SNN for Frame-based Event Vision Sensor</title>
      <link>https://iris-lab.skku.edu/publication/c89_icce_asia_2025/</link>
      <pubDate>Mon, 08 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c89_icce_asia_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C88] Continual Test-Time Fine-Tuning of Frame-Based Style Transfer Network for Video Stream Data</title>
      <link>https://iris-lab.skku.edu/publication/c88_vcip_2025/</link>
      <pubDate>Sun, 07 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c88_vcip_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C87] Perturb a Model, Not an Image: Towards Robust Privacy Protection via Anti-Personalized Diffusion Models</title>
      <link>https://iris-lab.skku.edu/publication/c87_neurips_2025/</link>
      <pubDate>Sat, 06 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c87_neurips_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C86] Efficient Multi-bit Quantization Network Training via Weight Bias Correction and Bit-wise Coreset Sampling</title>
      <link>https://iris-lab.skku.edu/publication/c86_neurips_2025/</link>
      <pubDate>Thu, 04 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c86_neurips_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C85] Optimized Minimal 3D Gaussian Splatting</title>
      <link>https://iris-lab.skku.edu/publication/c85_neurips_2025/</link>
      <pubDate>Wed, 03 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c85_neurips_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C84] Learnable Center-Based Quantization for Efficient Analog PIM with Reduced ADC Precision</title>
      <link>https://iris-lab.skku.edu/publication/c84_asp_dac_2026/</link>
      <pubDate>Tue, 02 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c84_asp_dac_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Prof. Bokyung Kim’s Seminar on ‘DPIMA: A DRAM-Based Processing-in-Memory Accelerator for Privacy-Preserving Machine Learning&#39;</title>
      <link>https://iris-lab.skku.edu/post/seminar_bokyung_kim/</link>
      <pubDate>Tue, 02 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/seminar_bokyung_kim/</guid>
      <description>&lt;p&gt;2025년 9월 2일, Rutgers University의 김보경 교수님이 성균관대학교를 방문해 “DPIMA: A DRAM-Based Processing-in-Memory Accelerator for Privacy-Preserving Machine Learning”이라는 주제로 세미나를 진행해 주셨습니다. 세미나는 반도체관 400321호에서 열렸고, IRIS 연구실의 고종환 교수님이 호스트를 맡았습니다.&lt;/p&gt;
&lt;p&gt;이번 발표는 ISLPED 2025에서 Best Paper로 선정된 해당 논문을 소개해 주신 자리였습니다. 교수님께서는 프라이버시 보존형 머신러닝을 보다 효율적으로 지원하기 위한 처리-인-메모리(PIM) 관점의 접근을 설명하시며, DRAM의 병렬성을 활용해 데이터 이동을 줄이고 학습 과정에 맞춘 효율적 데이터 흐름을 적용하는 큰 그림을 공유해 주셨습니다. 연구의 의의와 실제 적용 가능성에 대한 논의도 이어져, 프라이버시 보존형 학습을 위한 하드웨어·시스템적 관점에 대해 한층 깊이 있는 이해가 가능했습니다.&lt;/p&gt;
&lt;p&gt;간단한 약력으로, 김보경 교수님은 Rutgers University 전기·컴퓨터공학과에서 재직 중이며, 메모리 중심 가속기와 PIM, 혼성신호 VLSI, 신뢰할 수 있고 효율적인 머신러닝용 칩·아키텍처·시스템을 연구하고 있습니다. 박사학위는 Duke University에서 취득했습니다.&lt;/p&gt;
&lt;p&gt;바쁜 일정에도 귀한 시간을 내어 주신 김보경 교수님께 깊이 감사드립니다.&lt;/p&gt;
&lt;p&gt;On September 2, 2025, Prof. Bokyung Kim from Rutgers University visited Sungkyunkwan University and delivered a seminar titled “DPIMA: A DRAM-Based Processing-in-Memory Accelerator for Privacy-Preserving Machine Learning.” The seminar took place in the Semiconductor Building, Room 400321, and was hosted by Prof. Jong Hwan Ko for the IRIS Lab.&lt;/p&gt;
&lt;p&gt;This talk introduced the paper that was selected as the ISLPED 2025 Best Paper. Prof. Kim outlined a processing-in-memory perspective for enabling privacy-preserving machine learning more efficiently, describing how leveraging DRAM-level parallelism can reduce data movement and how training-phase–aware dataflows can be applied in practice. The discussion highlighted the significance of the work and its potential for real-world deployment, offering the audience a deeper understanding of hardware and system approaches to privacy-preserving training.&lt;/p&gt;
&lt;p&gt;By way of a brief bio, Prof. Kim is a faculty member in Electrical and Computer Engineering at Rutgers University. Her research spans memory-centric accelerators and PIM, mixed-signal VLSI, and trustworthy, efficient machine-learning chips, architectures, and systems. She received her Ph.D. from Duke University.&lt;/p&gt;
&lt;p&gt;We sincerely thank Prof. Bokyung Kim for sharing her time and insights with the IRIS Lab.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C83] Data Flow-Aware Weight Remapping for Efficient Fault Tolerance in ReRAM-Based Accelerators</title>
      <link>https://iris-lab.skku.edu/publication/c83_asp_dac_2026/</link>
      <pubDate>Mon, 01 Sep 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c83_asp_dac_2026/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C82] ETA: Efficient Transformer Attention Mapping for ReRAM-based Compute-In-Memory Architectures</title>
      <link>https://iris-lab.skku.edu/publication/c82_apccas_2025/</link>
      <pubDate>Sun, 03 Aug 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c82_apccas_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C81] Weight Sharing for Array-Efficient CNN Inference in Compute-In-Memory Architectures</title>
      <link>https://iris-lab.skku.edu/publication/c81_isocc_2025/</link>
      <pubDate>Sat, 02 Aug 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c81_isocc_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C80] Efficient and Robust SNN Decoder Training for Closed-Loop Brain-Machine Interfaces</title>
      <link>https://iris-lab.skku.edu/publication/c80_biocas_2025/</link>
      <pubDate>Fri, 01 Aug 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c80_biocas_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>MIT Technology Review Features IRIS Lab’s Work on Voice Unlearning for Zero-Shot TTS</title>
      <link>https://iris-lab.skku.edu/post/interview_mit_tech/</link>
      <pubDate>Tue, 15 Jul 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/interview_mit_tech/</guid>
      <description>&lt;p&gt;최근 &lt;em&gt;MIT Technology Review&lt;/em&gt;에 고종환 교수님과 김진주 석사과정 연구원의 인터뷰가 실렸습니다 ( &lt;a href=&#34;https://www.technologyreview.com/2025/07/15/1120094/ai-text-to-speech-programs-could-one-day-unlearn/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.technologyreview.com/2025/07/15/1120094/ai-text-to-speech-programs-could-one-day-unlearn/&lt;/a&gt; ). 이번 인터뷰는 피터 홀(Peter Hall) 편집자와 함께 최근 ICML 2025에서 발표한 논문 *“Do Not Mimic My Voice: Speaker Identity Unlearning for Zero-Shot Text-to-Speech”*을 중심으로 진행되었는데요. 기사에서는 AI 음성 합성(Text-to-Speech) 시스템이 특정 화자의 목소리 모방 능력을 “잊을(unlearn)” 수 있는 가능성과, 이를 통해 오디오 딥페이크로부터 사용자를 보호할 수 있는 새로운 방향을 다루고 있습니다.&lt;/p&gt;
&lt;p&gt;인터뷰에서는 Meta의 Voicebox와 같은 최신 zero-shot TTS 모델이 단 몇 초의 음성만으로도 화자의 목소리를 그대로 복제할 수 있다는 점을 지적하며, 이로 인한 프라이버시와 윤리적 문제를 짚었습니다. 저희 논문은 이런 문제를 해결하기 위해 특정 화자의 목소리만 선택적으로 제거하면서도 다른 화자에 대한 성능은 유지할 수 있는 Teacher-Guided Unlearning (TGU) 기법을 제안했습니다. 또 모델이 실제로 화자를 얼마나 잘 “잊었는지” 평가할 수 있도록 speaker-Zero Retrain Forgetting (spk-ZRF)이라는 새로운 지표를 도입했으며, 실험을 통해 특정 화자의 목소리 복제 능력을 크게 줄이면서도 전체 음성 품질은 유지할 수 있음을 확인했습니다.&lt;/p&gt;
&lt;p&gt;저희 연구는 단순히 기술적 호기심을 넘어서, 앞으로는 사용자가 “내 목소리를 복제하지 마라(Do not mimic my voice)”라고 요구하면 시스템 차원에서 이를 존중할 수 있는 미래를 열 수 있음을 보여줍니다. 목소리 복제와 오디오 딥페이크 기술이 점점 정교해지는 지금, 이러한 기능은 필수적인 프라이버시 보호 장치가 될 수 있음을 인터뷰에서 강조합니다.&lt;/p&gt;
&lt;p&gt;Recently, &lt;em&gt;MIT Technology Review&lt;/em&gt; published an interview with Prof. Jong Hwan Ko and M.S. student Jinju Kim ( &lt;a href=&#34;https://www.technologyreview.com/2025/07/15/1120094/ai-text-to-speech-programs-could-one-day-unlearn/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.technologyreview.com/2025/07/15/1120094/ai-text-to-speech-programs-could-one-day-unlearn/&lt;/a&gt; ). The interview with editor Peter Hall focused on our ICML 2025 paper, &lt;em&gt;“Do Not Mimic My Voice: Speaker Identity Unlearning for Zero-Shot Text-to-Speech.”&lt;/em&gt; The article explores the possibility that AI text-to-speech systems could “unlearn” their ability to imitate specific speakers, opening a new direction for protecting users from audio deepfakes.&lt;/p&gt;
&lt;p&gt;In the interview, we note that state-of-the-art zero-shot TTS models such as Meta’s Voicebox can replicate a speaker’s voice from just a few seconds of audio, raising serious privacy and ethical concerns. Our paper proposes Teacher-Guided Unlearning (TGU), a technique that selectively removes a targeted speaker’s voice while maintaining performance on others. We also introduce speaker-Zero Retrain Forgetting (spk-ZRF) to quantify how well the model “forgets,” and we show experimentally that targeted voice cloning can be greatly reduced without degrading overall speech quality.&lt;/p&gt;
&lt;p&gt;Our research goes beyond technical curiosity, pointing toward a future in which users can say “Do not mimic my voice” and have that respected at the system level. As voice cloning and audio deepfakes become increasingly sophisticated, this capability can serve as an essential safeguard for voice privacy.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Joins Global Basic Research Laboratory (BRL)</title>
      <link>https://iris-lab.skku.edu/post/brl/</link>
      <pubDate>Fri, 11 Jul 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/brl/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님이 경희대학교 컴퓨터공학부 김휘용 교수님이 총괄하는 &amp;lsquo;2025년도 글로벌 기초연구실 지원사업(BRL, Basic Research Laboratory)&amp;rsquo; 과제에 공동연구진으로 참여하게 되었습니다.&lt;/p&gt;
&lt;p&gt;본 사업은 한국연구재단이 주관하는 글로벌 기초연구실 지원사업으로, 2025년 6월부터 3년간 총 15억 원 규모로 진행됩니다. 연구 주제는 &amp;lsquo;인간과 기계 시각을 동시에 지원하는 다목적 시각정보 압축 연구&amp;rsquo;로, 공동연구진으로는 배성호 교수님, 최진우 교수님, 고종환 교수님이 참여합니다.&lt;/p&gt;
&lt;p&gt;이미지와 비디오와 같은 시각정보는 인간이 콘텐츠로 소비하거나 기계(인공지능)가 추론을 위해 소비하는 등 그 활용법이 점점 증가하고 있으나, 시각정보의 특성상 데이터량이 방대하여 효과적인 압축 기술이 반드시 필요합니다. 그러나 기존의 시각정보 압축 연구는 인간시각과 기계시각 각각만을 위해 최적화되거나 다양한 센서로부터 획득된 이종 시각정보를 처리하는 데 한계가 있습니다.&lt;/p&gt;
&lt;p&gt;연구팀은 이러한 문제를 해결하기 위해 신경망 기반의 통합 프레임워크를 제안하였으며, 향후 3년간 다양한 시각정보(가시광, 초분광, SAR 등)와 다양한 목적(인간시각과 기계분석)을 동시에 지원하기 위한 원천기술 확보 및 국제 표준화를 추진할 예정입니다.&lt;/p&gt;
&lt;p&gt;IRIS 연구실은 이 과제에서 에너지 효율적이고 적응적인 다목적 부호화 기술을 담당합니다. 제한된 연산 자원과 전력 환경에서도 인간 시각과 기계 분석을 동시에 지원할 수 있는 압축 및 부호화 방법론을 개발하며, 하드웨어-소프트웨어 공동 최적화를 통해 실용적인 솔루션을 제시할 계획입니다. 이는 IRIS 연구실이 그동안 축적해 온 자원 효율적 AI 기술과 멀티모달 처리 전문성이 시각정보 압축 분야에 적용되는 중요한 기회입니다.&lt;/p&gt;
&lt;p&gt;연구팀은 중국 저장대학교의 Lu Yu 교수팀, 미국 샌디에고 대학교의 Hao Su 교수팀, 캐나다 사이먼 프레이저 대학교의 Ivan V. Bajic 교수팀, 오스트리아 클라겐푸르트 대학교의 Hadi Amirpour 교수팀, Intel Labs의 Kyle Min 박사팀과 공동연구 MoU 체결을 완료하는 등, 활발한 국제 협력을 통해 글로벌 기술 선도 및 국제표준화에 박차를 기울 예정입니다.&lt;/p&gt;
&lt;p&gt;본 과제를 통해 IRIS 연구실은 시각정보 압축 분야에서 에너지 효율성과 적응성을 결합한 새로운 연구 방향을 개척하고, 글로벌 협력 네트워크를 확장하며, 차세대 멀티모달 압축 기술의 실용화에 기여할 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;자세한 내용은 경희대학교 컴퓨터공학과 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ce.khu.ac.kr/ce/user/bbs/BMSR00040/list.do?menuNo=1600120&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ce.khu.ac.kr/ce/user/bbs/BMSR00040/list.do?menuNo=1600120&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab&amp;rsquo;s Prof. Jong Hwan Ko will participate as a co-researcher in the &amp;lsquo;2025 Global Basic Research Laboratory (BRL)&amp;rsquo; project led by Prof. Kim Hwiyong from Kyung Hee University&amp;rsquo;s Department of Computer Engineering.&lt;/p&gt;
&lt;p&gt;This project is supported by the National Research Foundation of Korea with a budget of 1.5 billion KRW over three years starting from June 2025. The research theme is &amp;ldquo;Multipurpose Visual Information Compression Supporting Both Human and Machine Vision,&amp;rdquo; with co-researchers including Prof. Seongho Bae, Prof. Jinwoo Choi, and Prof. Jong Hwan Ko.&lt;/p&gt;
&lt;p&gt;Visual information such as images and videos is increasingly consumed by humans as content or by machines (AI) for inference, but the massive data volume inherent to visual information necessitates effective compression technologies. However, existing visual information compression research has limitations in being optimized only for human vision or machine vision separately, or in processing heterogeneous visual information acquired from various sensors.&lt;/p&gt;
&lt;p&gt;The research team proposes a unified neural network-based framework to address these challenges, aiming to secure fundamental technologies and promote international standardization over the next three years to simultaneously support various visual information types (visible light, hyperspectral, SAR, etc.) and various purposes (human vision and machine analysis).&lt;/p&gt;
&lt;p&gt;IRIS Lab is responsible for energy-efficient and adaptive multipurpose coding technologies in this project. We will develop compression and coding methodologies that can simultaneously support human vision and machine analysis even in resource-constrained and power-limited environments, presenting practical solutions through hardware-software co-optimization. This represents an important opportunity to apply IRIS Lab&amp;rsquo;s accumulated expertise in resource-efficient AI and multimodal processing to visual information compression.&lt;/p&gt;
&lt;p&gt;The research team has completed MoU agreements for joint research with Prof. Lu Yu&amp;rsquo;s team at Zhejiang University (China), Prof. Hao Su&amp;rsquo;s team at UC San Diego (USA), Prof. Ivan V. Bajic&amp;rsquo;s team at Simon Fraser University (Canada), Prof. Hadi Amirpour&amp;rsquo;s team at Klagenfurt University (Austria), and Dr. Kyle Min&amp;rsquo;s team at Intel Labs, accelerating global technology leadership and international standardization through active international collaboration.&lt;/p&gt;
&lt;p&gt;Through this project, IRIS Lab expects to pioneer new research directions combining energy efficiency and adaptability in visual information compression, expand its global collaboration network, and contribute to the practical implementation of next-generation multimodal compression technologies.&lt;/p&gt;
&lt;p&gt;For more details, please visit the official news from Kyung Hee University&amp;rsquo;s Department of Computer Engineering: &lt;a href=&#34;https://ce.khu.ac.kr/ce/user/bbs/BMSR00040/list.do?menuNo=1600120&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ce.khu.ac.kr/ce/user/bbs/BMSR00040/list.do?menuNo=1600120&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C79] Row-Column Hybrid Grouping for Fault-Resilient Multi-Bit Weight Representation on IMC Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c79_iccad_2025/</link>
      <pubDate>Tue, 01 Jul 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c79_iccad_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C78] MSQ: Memory-Efficient Bit Sparsification Quantization</title>
      <link>https://iris-lab.skku.edu/publication/c78_iccv_2025/</link>
      <pubDate>Thu, 26 Jun 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c78_iccv_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Featured on KBS National News</title>
      <link>https://iris-lab.skku.edu/post/youtube_kbs/</link>
      <pubDate>Tue, 17 Jun 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/youtube_kbs/</guid>
      <description>&lt;p&gt;IRIS 연구실이 AI 연구 개발 현장의 현실을 조명한 KBS 뉴스 연속 기획에 소개되었습니다. (&lt;a href=&#34;https://www.youtube.com/watch?v=fNRObV4AqqU&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.youtube.com/watch?v=fNRObV4AqqU&lt;/a&gt;)&lt;/p&gt;
&lt;p&gt;2025년 6월 17일 방송된 KBS 뉴스는 &amp;ldquo;AI 3대 강국을 향한 도전&amp;quot;이라는 주제로 국내 AI 연구 현장의 전력 인프라 문제를 다뤘습니다. 우리 연구실 소속 대학원생이 직접 인터뷰에 참여하여, GPU 서버 자원의 효율적 활용과 연구 현장의 현실을 생생하게 전달했습니다. 또한 고종환 교수님께서는 제한된 자원 환경이 AI 연구의 혁신과 국제 경쟁력에 미치는 영향에 대해 전문가 의견을 제시하셨습니다.&lt;/p&gt;
&lt;p&gt;이번 보도는 단순히 어려움을 지적하는 것을 넘어, 한국 AI 연구의 미래를 위해 필요한 인프라 투자와 정책적 지원의 중요성을 사회적으로 환기시키는 계기가 되었습니다. 우리 연구실은 제한된 자원 속에서도 효율적인 GPU 관리 시스템을 운영하며, 세계 수준의 AI 연구 성과를 만들어가고 있습니다.&lt;/p&gt;
&lt;p&gt;앞으로도 IRIS 연구실은 더 나은 연구 환경을 만들어가며, AI 기술 발전에 기여하는 혁신적인 연구를 이어갈 것입니다.&lt;/p&gt;
&lt;p&gt;IRIS Lab was featured in a KBS News investigative series highlighting the realities of AI research and development in Korea. (&lt;a href=&#34;https://www.youtube.com/watch?v=fNRObV4AqqU&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.youtube.com/watch?v=fNRObV4AqqU&lt;/a&gt;)&lt;/p&gt;
&lt;p&gt;The June 17, 2025 KBS News broadcast focused on &amp;ldquo;The Challenge Toward Becoming a Top 3 AI Nation,&amp;rdquo; examining power infrastructure challenges facing domestic AI research. Our graduate student participated in an on-camera interview, providing firsthand insights into efficient GPU server resource management and the day-to-day realities of research work. Professor Jong Hwan Ko also contributed expert commentary on how limited resources impact AI research innovation and international competitiveness.&lt;/p&gt;
&lt;p&gt;This coverage went beyond simply pointing out difficulties—it served as a catalyst for raising public awareness about the critical need for infrastructure investment and policy support for Korea&amp;rsquo;s AI research future. Despite resource constraints, our lab operates efficient GPU management systems and continues producing world-class AI research outcomes.&lt;/p&gt;
&lt;p&gt;Moving forward, IRIS Lab will continue building better research environments and pursuing innovative research that advances AI technology.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Joins AI Semiconductor Innovation Institute for Physical AI Research</title>
      <link>https://iris-lab.skku.edu/post/semiconductor/</link>
      <pubDate>Mon, 09 Jun 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/semiconductor/</guid>
      <description>&lt;p&gt;IRIS 연구실의 고종환 교수님이 성균관대학교 &amp;lsquo;AI반도체 혁신연구소&amp;rsquo;의 Physical AI 센터에 참여하게 되었습니다.&lt;/p&gt;
&lt;p&gt;AI반도체 혁신연구소는 과학기술정보통신부 주관 &amp;lsquo;2025년 산학연계 AI반도체 선도기술 인재양성 사업&amp;rsquo;의 일환으로 설립되며, 향후 5년 6개월간 총 110억 원의 정부 출연금을 지원받아 운영됩니다. 본 사업은 AI반도체 분야 실전형 고급 인재를 양성하기 위한 정부 전략사업으로, 성균관대는 삼성전자, 모빌린트, 보스반도체, 오픈엣지테크놀로지 등 온디바이스 AI 선도 기업들과 긴밀히 협력합니다.&lt;/p&gt;
&lt;p&gt;고종환 교수님은 Physical AI 센터에서 다중센서 멀티모달 융합 처리를 위한 하드웨어/소프트웨어 최적화 연구를 담당합니다. Physical AI는 자율주행 차량, 로봇, 드론 등 물리적 환경과 실시간으로 상호작용하는 AI 시스템의 핵심 기술입니다. 이러한 시스템들은 카메라, 라이다, 레이더, IMU 등 다양한 센서로부터 실시간으로 들어오는 이기종 데이터를 효율적으로 융합하고 처리해야 합니다.&lt;/p&gt;
&lt;p&gt;IRIS 연구실은 이 과제를 통해 멀티모달 데이터 융합 과정에서 하드웨어 자원을 효율적으로 활용하고 소프트웨어 알고리즘을 최적화하는 연구를 수행합니다. 특히 제한된 전력과 연산 능력을 가진 온디바이스 환경에서 고성능 인식과 판단을 가능하게 하는 것이 핵심 과제입니다. 클라우드에 의존하지 않고 즉각적인 반응이 필요한 자율주행과 로보틱스 분야에서 이러한 기술은 매우 중요합니다.&lt;/p&gt;
&lt;p&gt;연구실은 그동안 축적해 온 하드웨어-소프트웨어 공동 설계 전문성을 바탕으로, 센서 데이터의 특성에 맞는 최적화된 처리 구조를 개발하고, 각 센서 모달리티 간의 효율적인 융합 메커니즘을 연구합니다. 이는 단순히 알고리즘 개발에 그치지 않고, 실제 하드웨어 제약을 고려한 실용적인 솔루션을 제시하는 것을 목표로 합니다.&lt;/p&gt;
&lt;p&gt;본 사업은 연간 약 60명의 석박사과정 학생들이 참여 기업과의 공동 프로젝트에 직접 참여하는 실무 중심 프로그램입니다. IRIS 연구실 학생들도 이 과정을 통해 Physical AI 분야의 최신 산업 기술과 연구 역량을 동시에 쌓을 수 있으며, 삼성전자를 비롯한 국내 대표 반도체 기업들과 협력하며 실전 경험을 얻게 됩니다.&lt;/p&gt;
&lt;p&gt;이번 참여를 통해 IRIS 연구실은 Physical AI 분야에서 멀티모달 센서 융합과 하드웨어-소프트웨어 공동 최적화 기술을 한층 발전시키고, 온디바이스 AI 시스템의 자원 효율성을 극대화하는 연구 성과를 창출할 계획입니다. 또한 산업체와의 긴밀한 협력을 통해 연구실의 기술이 실제 제품과 서비스로 구현되는 기회를 확대할 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;자세한 내용은 정보통신대학 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=202113&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=202113&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab&amp;rsquo;s Prof. Jong Hwan Ko will participate in the Physical AI Center of SKKU&amp;rsquo;s AI Semiconductor Innovation Institute.&lt;/p&gt;
&lt;p&gt;The AI Semiconductor Innovation Institute is established as part of the Ministry of Science and ICT&amp;rsquo;s &amp;lsquo;2025 Industry-Academia Collaboration AI Semiconductor Leading Technology Talent Development Program,&amp;rsquo; receiving 11 billion KRW over five and a half years. This government strategic initiative aims to cultivate advanced professionals in AI semiconductors, with SKKU collaborating closely with leading on-device AI companies including Samsung Electronics, Mobilint, Voss Semiconductor, and OpenEdge Technology.&lt;/p&gt;
&lt;p&gt;Prof. Jong Hwan Ko will focus on hardware/software optimization for multi-sensor multimodal fusion processing in the Physical AI Center. Physical AI is a core technology for AI systems that interact in real-time with physical environments, such as autonomous vehicles, robots, and drones. These systems must efficiently fuse and process heterogeneous data streaming in real-time from diverse sensors like cameras, LiDAR, radar, and IMUs.&lt;/p&gt;
&lt;p&gt;Through this project, IRIS Lab will research how to efficiently utilize hardware resources and optimize software algorithms in the multimodal data fusion process. The&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C77] Multi-Frame ISP: Enhancing Vision-Based Tasks with RAW and Infrared Videos</title>
      <link>https://iris-lab.skku.edu/publication/c77_avss_2025/</link>
      <pubDate>Mon, 02 Jun 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c77_avss_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C76] TruncQuant: Truncation-Ready Quantization for DNNs with Flexible Weight Bit Precision</title>
      <link>https://iris-lab.skku.edu/publication/c76_islped_2025/</link>
      <pubDate>Sun, 01 Jun 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c76_islped_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C75] Do Not Mimic My Voice : Speaker Identity Unlearning for Zero-Shot Text-to-Speech</title>
      <link>https://iris-lab.skku.edu/publication/c75_icml_2025/</link>
      <pubDate>Thu, 01 May 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c75_icml_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J34] Mixed Binary Supporting Electrolyte Approach for Enhanced Synaptic Functionality in One-shot Integrable Electropolymerized Synaptic Transistors</title>
      <link>https://iris-lab.skku.edu/publication/j34_materials_2025/</link>
      <pubDate>Wed, 23 Apr 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j34_materials_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J33] GAROS: Genetic Algorithm-Aided Row-Skipping for Shift and Duplicate Kernel Mapping in Processing-In-Memory Architectures</title>
      <link>https://iris-lab.skku.edu/publication/j33_jsa_2025/</link>
      <pubDate>Tue, 01 Apr 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j33_jsa_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C74] Test-Time Fine-Tuning of Image Compression Models for Multi-Task Adaptability</title>
      <link>https://iris-lab.skku.edu/publication/c74_cvpr_2025/</link>
      <pubDate>Sat, 01 Mar 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c74_cvpr_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title> City University of Hong Kong (CityU) Collaboration (Visiting Researcher: Chanwook Hwang)</title>
      <link>https://iris-lab.skku.edu/post/visiting_researcher_chanwook_hwang/</link>
      <pubDate>Tue, 14 Jan 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/visiting_researcher_chanwook_hwang/</guid>
      <description>&lt;p&gt;IRIS 연구실 석박통합과정 황찬욱 연구원은 2024년 8월부터 2025년 1월까지 홍콩시립대학교 (City University of Hong Kong, 이하 CityU) BRAINSys Lab에서 방문 연구원으로 활동하며, 뇌-기계 인터페이스(Brain-Machine Interface, BMI)와 뉴로모픽 회로 분야의 세계적 전문가인 Arindam Basu 교수님과 함께 차세대 신경 신호 압축 및 스파이킹 신경망 기반 신호처리에 관한 연구를 수행하였습니다.&lt;/p&gt;
&lt;p&gt;이번 방문 연구는 GVIC3 (Global Value Innovation Creator) 사업의 지원을 받아 진행되었으며, 성균관대학교 IRIS 연구실과 CityU BRAINSys Lab 간의 공동연구 교류를 촉진하였습니다. 이를 통해 황찬욱 연구원은 실험실 연구 문화 경험과 국제 세미나 참여를 통해 글로벌 연구 네트워크를 확장하는 중요한 계기를 마련했습니다.&lt;/p&gt;
&lt;p&gt;특히 본 연구에서 황찬욱 연구원은 이식형 BMI(implantable BMI)를 위한 신경 신호 압축(Neural Signal Compression) 및 이벤트 기반 스파이크 검출(Event-based Spike Detector, SPD) 문제를 다루며, 이를 스파이킹 신경망(Spiking Neural Network, SNN)으로 구현한 SNN-SPD 알고리즘을 제안하였습니다. 그 결과, 연산량을 99% 감소시키고, 모델 크기를 73% 줄이면서도 정확도를 2% 향상시키는 성과를 달성했습니다.&lt;/p&gt;
&lt;p&gt;이 성과는 2025년 5월, 영국 런던에서 개최된 IEEE ISCAS (International Symposium on Circuits and Systems) 학회에서 발표되었으며, 차세대 이식형 BMI와 뉴로모픽 신호처리 연구의 국제적 주목을 받았습니다.&lt;/p&gt;
&lt;p&gt;방문 연구 종료 후에도, 황찬욱 연구원은 본 연구 성과를 바탕으로 차세대 이식형 BMI에서 요구되는 초저전력·고효율 신호처리 회로 설계 연구를 확장할 계획입니다. 또한 CityU에서의 경험을 통해 구축된 국제 공동연구 네트워크를 활용하여, 뇌신호 인터페이스 및 뉴로모픽 하드웨어 시스템 분야에서 지속적인 협력 연구를 이어갈 예정입니다.&lt;/p&gt;
&lt;p&gt;황찬욱 연구원의 CityU 방문은 한국과 홍콩 간의 연구 교류를 넘어, 스파이킹 신경망·이식형 BMI·뉴로모픽 신호처리 분야에서 새로운 국제 공동연구의 시너지를 창출할 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;From August 2024 to January 2025, Chanwook Hwang, a Combined M.S. and Ph.D. student at IRIS Lab, served as a visiting researcher at the BRAINSys Lab, City University of Hong Kong (CityU). During his stay, he collaborated with Prof. Arindam Basu, a leading expert in neuromorphic circuits and brain–machine interfaces (BMI), to advance research on neural signal compression and spiking neural network (SNN)-based spike detection.&lt;/p&gt;
&lt;p&gt;This research visit was supported by the GVIC3 (Global Value Innovation Creator) program, which fostered collaborative exchange between SKKU IRIS Lab and CityU BRAINSys Lab. Through this program, Chanwook Hwang broadened his academic network by engaging in international seminars and immersing himself in CityU’s lab culture.&lt;/p&gt;
&lt;p&gt;In this project, Chanwook Hwang proposed the SNN-based Spike Detector (SNN-SPD) for next-generation implantable BMI systems. The method successfully achieved a 99% reduction in computation and 73% reduction in model size, while improving accuracy by 2%.&lt;/p&gt;
&lt;p&gt;This achievement was further recognized when this work was presented at IEEE ISCAS 2025 (International Symposium on Circuits and Systems), held in May 2025 in London, UK, where it attracted attention in the fields of implantable BMIs and neuromorphic signal processing.&lt;/p&gt;
&lt;p&gt;Following the completion of his visit, Chanwook Hwang plans to extend this work toward ultra-low-power and high-efficiency neural signal processing circuits essential for implantable BMI systems. Furthermore, leveraging the international research network established during his stay, he will continue collaborative efforts in the domains of neural interfaces and neuromorphic hardware systems.&lt;/p&gt;
&lt;p&gt;Chanwook Hwang’s visit to CityU represents not only a bridge between Korea and Hong Kong in advanced research but also a step forward in fostering global collaboration in spiking neural networks, implantable BMI, and neuromorphic signal processing.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C73] Event-based Neural Spike Detection Using Spiking Neural Networks for Neuromorphic iBMI System</title>
      <link>https://iris-lab.skku.edu/publication/c73_iscas_2025/</link>
      <pubDate>Wed, 01 Jan 2025 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c73_iscas_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J32] Input/Mapping Precision Controllable Digital CIM with Adaptive Adder Tree Architecture for Flexible DNN Inference</title>
      <link>https://iris-lab.skku.edu/publication/j32_jsa_2025/</link>
      <pubDate>Sun, 01 Dec 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j32_jsa_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C72] MEMHD: Memory-Efficient Multi-Centroid Hyperdimensional Computing for Fully-Utilized In-Memory Computing Architectures</title>
      <link>https://iris-lab.skku.edu/publication/c72_date_2025/</link>
      <pubDate>Wed, 13 Nov 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c72_date_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C71] Column-wise Quantization of Weights and Partial Sums for Accurate and Efficient Compute-In-Memory Accelerators</title>
      <link>https://iris-lab.skku.edu/publication/c71_date_2025/</link>
      <pubDate>Tue, 12 Nov 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c71_date_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C70] Low-Rank Compression for IMC Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c70_date_2025/</link>
      <pubDate>Mon, 11 Nov 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c70_date_2025/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C69] Fully Approximate Computing for Efficient Multi-bit DNN Inference in CIM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c69_icce_asia_2024/</link>
      <pubDate>Fri, 01 Nov 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c69_icce_asia_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C68] Energy-Efficient Video Streaming: A Study on Bit Depth and Color Subsampling</title>
      <link>https://iris-lab.skku.edu/publication/c68_ieee_vcip_2024/</link>
      <pubDate>Fri, 27 Sep 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c68_ieee_vcip_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J31] An FPGA-Based Energy-Efficient Real-Time Hand Pose Estimation System with an Integrated Image Signal Processor for Indirect 3D Time-of-Flight Sensors</title>
      <link>https://iris-lab.skku.edu/publication/j31_ieee_iot_2024/</link>
      <pubDate>Sun, 01 Sep 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j31_ieee_iot_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C67] F-3DGS: Factorized Coordinates and Representations for 3D Gaussian Splatting</title>
      <link>https://iris-lab.skku.edu/publication/c67_acm_2024/</link>
      <pubDate>Sat, 20 Jul 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c67_acm_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C66] HandDAGT: A Denoising Adaptive Graph Transformer for 3D Hand Pose Estimation</title>
      <link>https://iris-lab.skku.edu/publication/c66_eccv_2024_hand/</link>
      <pubDate>Fri, 19 Jul 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c66_eccv_2024_hand/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Undergraduate Researchers Win Awards at IEIE Competition and URP Showcase</title>
      <link>https://iris-lab.skku.edu/post/ieie_award/</link>
      <pubDate>Thu, 18 Jul 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/ieie_award/</guid>
      <description>&lt;p&gt;IRIS 연구실 학부연구생들이 두 개의 경진대회에서 수상하는 성과를 거두었습니다.&lt;/p&gt;
&lt;p&gt;김진희, 윤서연 학생은 대한전자공학회 인공지능 학부생 논문 경진대회에서 최우수상을 수상했습니다. 이 대회는 학부생들의 연구 능력을 향상시키고 학문적 교류를 활성화하기 위해 대한전자공학회 하계학술대회에서 개최됩니다.&lt;/p&gt;
&lt;p&gt;두 학생은 하드웨어 자원에 따라 인공지능 모델의 정밀도를 유연하게 조절할 수 있는 동적신경망 연구를 수행했습니다. 기존 방식보다 메모리와 연산을 효율적으로 사용하면서도 다양한 하드웨어 환경에서 모델을 실행할 수 있도록 하는 기술을 개발했습니다.&lt;/p&gt;
&lt;p&gt;한편, 김진희, 강도영, 백승하 학생으로 구성된 학제간 팀은 융합연구학점제 성과보고회에서 성균융합원장상을 받았습니다. 융합연구학점제는 다양한 전공의 학생들이 함께 문제를 발굴하고 해법을 찾아가는 과정을 학점으로 인정하는 제도로, 이번 학기 처음 시행되었습니다.&lt;/p&gt;
&lt;p&gt;이 팀은 인공지능 학습 과정의 환경적 영향을 줄이기 위한 연구를 수행했습니다. 동적신경망 학습을 최적화하여 정확도를 유지하면서도 전체 학습시간을 단축시킬 수 있는 방법을 제시했으며, 이는 AI 학습의 에너지 소비를 줄이는 데 기여할 수 있습니다.&lt;/p&gt;
&lt;p&gt;이번 수상은 연구실에서 학부생들과 함께 만들어가는 연구 문화의 한 성과입니다. 학생들이 주도적으로 질문을 던지고 답을 찾아가는 과정에서 얻은 경험들이, 이들의 연구자로서의 여정에 의미 있는 출발점이 되기를 바랍니다.&lt;/p&gt;
&lt;p&gt;자세한 내용은 정보통신대학 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=178043&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=178043&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab undergraduate researchers won awards at two competitions.&lt;/p&gt;
&lt;p&gt;Jinhee Kim and Seoyeon Yoon won the Best Paper Award at the IEIE AI Undergraduate Paper Competition. This competition is held at the IEIE Summer Conference to enhance undergraduate research capabilities and activate academic exchange.&lt;/p&gt;
&lt;p&gt;The two students conducted research on dynamic neural networks that flexibly adjust AI model precision according to hardware resources. They developed technology that uses memory and computation more efficiently than existing methods while enabling model execution across diverse hardware environments.&lt;/p&gt;
&lt;p&gt;Meanwhile, an interdisciplinary team of Jinhee Kim, Doyeong Kang (Mathematics), and Seungha Baek (Mechanical Engineering) received the Director&amp;rsquo;s Award at the URP Showcase. The Undergraduate Research Program is a credit-recognized system where students from various majors work together to identify problems and find solutions, launching for the first time this semester.&lt;/p&gt;
&lt;p&gt;This team conducted research to reduce the environmental impact of AI training processes. They proposed methods to optimize dynamic neural network training, maintaining accuracy while shortening overall training time, which can contribute to reducing energy consumption in AI training.&lt;/p&gt;
&lt;p&gt;These awards represent an outcome of the research culture we build together with undergraduate students at the lab. We hope the experiences gained through the process of actively raising questions and finding answers will serve as a meaningful starting point in their journey as researchers.&lt;/p&gt;
&lt;p&gt;For more details, please visit the official news from the College of Information &amp;amp; Communication Engineering: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=178043&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=178043&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C65] Continuous Memory Representation for Anomaly Detection</title>
      <link>https://iris-lab.skku.edu/publication/c65_eccv_2024_memory/</link>
      <pubDate>Fri, 05 Jul 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c65_eccv_2024_memory/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J30] Reconfigurable Resistive Switching Memory for Telegraph Code Sensing and Recognizing Reservoir Computing Systems</title>
      <link>https://iris-lab.skku.edu/publication/j30_small_2024/</link>
      <pubDate>Wed, 12 Jun 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j30_small_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Featured on the Ministry of Science and ICT’s YouTube Channel</title>
      <link>https://iris-lab.skku.edu/post/youtube_msit/</link>
      <pubDate>Mon, 10 Jun 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/youtube_msit/</guid>
      <description>&lt;p&gt;IRIS 연구실이 과학기술정보통신부 유튜브 시리즈 ‘요즘이랩’에 출연했습니다 (&lt;a href=&#34;https://www.youtube.com/watch?v=-7Mc6fKFByg%29&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.youtube.com/watch?v=-7Mc6fKFByg)&lt;/a&gt;. 영상 제목은 “대한민국 온디바이스 AI의 미래 책임진다! 성균관대학교 IRIS 연구실”이며, 콘텐츠 취지에 맞게 실험실을 직접 찾아와 연구 방향과 문화를 소개하는 형식으로 제작되었습니다. 영상에서는 하드웨어·소프트웨어 융합을 통해 온디바이스 AI를 선도하기 위한 IRIS 연구실의 비전이 강조되었습니다.&lt;/p&gt;
&lt;p&gt;특히 IRIS 연구실이 온디바이스에 최적화된 하드웨어와 소프트웨어 AI를 동시에 폭넓게 연구하고 있다는 점, 진행 중인 다양한 프로젝트와 연구 주제, 그리고 연구실 분위기와 채용·연구 환경에 대한 생생한 이야기가 담겨 있어 연구실을 한눈에 이해할 수 있는 시간이었습니다. 교수님과 학생들의 인터뷰를 통해 실제 연구 현장의 목소리와 워라밸 문화도 엿볼 수 있었습니다.&lt;/p&gt;
&lt;p&gt;끝으로, 이번 영상은 IRIS 연구실이 지향하는 학제간 협력과 연구분야를 학계와 산업계 전반에 알릴 수 있는 좋은 계기가 되었습니다. 소개에 도움을 주신 모든 분들께 감사드리며, 인터뷰와 촬영에 참여한 연구실 구성원들에게도 고마움을 전합니다.&lt;/p&gt;
&lt;p&gt;The IRIS Lab at Sungkyunkwan University was featured on the Ministry of Science and ICT’s YouTube series “Yojeum-i-Lab” (&lt;a href=&#34;https://www.youtube.com/watch?v=-7Mc6fKFByg%29&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.youtube.com/watch?v=-7Mc6fKFByg)&lt;/a&gt;. Titled “Leading Korea’s On-Device AI Future! SKKU IRIS Lab,” the episode takes an on-site look at our lab’s research directions and culture. It highlights IRIS’s vision to advance on-device AI through close hardware–software integration.&lt;/p&gt;
&lt;p&gt;The video showcases our broad efforts across jointly optimized hardware and software for on-device AI, a range of ongoing projects and research topics, and candid perspectives on our lab culture, hiring, and research environment—providing a clear snapshot of who we are and what we do. Interviews with faculty and students also share day-to-day experiences and our emphasis on healthy work–life balance.&lt;/p&gt;
&lt;p&gt;Ultimately, this feature offered a valuable opportunity to share IRIS’s interdisciplinary collaboration and research domains spanning academia and industry. We appreciate everyone who supported the production, and we thank our lab members who participated in the filming and interviews.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C64] C-AFA: A Conditionally Approximate Full Adder for Efficient DNN Inference in CIM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c64_isocc_2024_cim/</link>
      <pubDate>Wed, 05 Jun 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c64_isocc_2024_cim/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C63] A Hybrid Precision Network with Efficient Processing Elements for 3D Hand Pose Estimation</title>
      <link>https://iris-lab.skku.edu/publication/c63_isocc_2024_hand/</link>
      <pubDate>Tue, 04 Jun 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c63_isocc_2024_hand/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C62] Row-Efficient Pruning for In-Memory Convolutional Weight Mapping</title>
      <link>https://iris-lab.skku.edu/publication/c62_isocc_2024_inmemory/</link>
      <pubDate>Thu, 30 May 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c62_isocc_2024_inmemory/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J29] One-Shot Remote Integration of Macromolecular Synaptic Elements on a Chip for Ultrathin Flexible Neural Network System</title>
      <link>https://iris-lab.skku.edu/publication/j29_adma_2024/</link>
      <pubDate>Wed, 22 May 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j29_adma_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>ICLR 2024 Spotlight &amp; CVPR 2024 Highlight: Efficient 3D Scene Representation</title>
      <link>https://iris-lab.skku.edu/post/iclr2024_cvpr2024_highlight/</link>
      <pubDate>Mon, 20 May 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/iclr2024_cvpr2024_highlight/</guid>
      <description>&lt;p&gt;IRIS 연구실의 이주찬 연구원과 노다니엘 연구원이 참여한 두 편의 논문이 각각 ICLR 2024 Spotlight와 CVPR 2024 Highlight에 선정되었습니다.&lt;/p&gt;
&lt;p&gt;본 연구는 고종환 교수님, 박은병 교수님님이 공동으로 지도한 프로젝트로, 복잡한 3차원 장면을 효율적으로 표현하는 두 가지 방법론을 제시합니다. 첫 번째는 뉴럴 네트워크와 그리드 방식을 융합한 Coordinate-Aware Modulation이며, 두 번째는 소형화된 3D 가우시안 표현 기법입니다.&lt;/p&gt;
&lt;p&gt;Coordinate-Aware Modulation 연구는 3차원 이미지나 비디오를 표현할 때 뉴럴 네트워크의 각 레이어마다 그리드의 특징 벡터를 모듈레이션 방식으로 융합했습니다. 기존 방식은 큰 용량을 필요로 했지만, 본 연구는 매우 소형의 그리드를 사용하여 고주파 신호를 효율적으로 표현했습니다. 이미지, 비디오, 3차원 모델 등 다양한 미디어 데이터에 적용한 결과 적은 네트워크 크기로도 우수한 신호 복원 능력을 보였습니다.&lt;/p&gt;
&lt;p&gt;Compact 3D Gaussian Splatting 연구는 3D 가우시안 형태로 장면을 표현하는 방식의 저장용량 문제를 해결했습니다. 최근 3D 가우시안 방식은 100 FPS 이상의 빠른 렌더링이 가능하지만 매우 큰 저장용량을 필요로 했습니다. 본 연구는 가우시안의 수를 렌더링 성능 감소 없이 줄이는 데 성공했으며, 새로운 표현 방법론을 제시하여 고성능과 효율적인 저장 공간을 동시에 달성했습니다. 실제 데이터셋 평가에서 렌더링 품질 저하 없이 25배 이상의 저장용량 감소와 렌더링 속도 향상을 이뤘습니다.&lt;/p&gt;
&lt;p&gt;첫 번째 연구는 기계학습 분야 최우수 학술대회인 ICLR 2024에서 제출 논문의 상위 6%에 해당하는 Spotlight에 선정되었으며, 두 번째 연구는 컴퓨터비전 분야 최우수 학술대회인 CVPR 2024에서 상위 3%에 해당하는 Highlight에 선정되었습니다. 이러한 효율적인 3D 표현 기술은 NeRF, 생성 모델, 메타버스 등 다양한 분야에서 실용적으로 활용될 수 있을 것으로 기대되며, IRIS 연구실은 이를 기반으로 더욱 발전된 3D 비전 기술을 연구해 나갈 계획입니다.&lt;/p&gt;
&lt;p&gt;논문 1: Coordinate-Aware Modulation for Neural Fields&lt;br&gt;
연구 홈페이지: &lt;a href=&#34;https://maincold2.github.io/cam/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://maincold2.github.io/cam/&lt;/a&gt;&lt;br&gt;
저자: 이주찬, 노다니엘, 남승태, 고종환, 박은병&lt;/p&gt;
&lt;p&gt;논문 2: Compact 3D Gaussian Representation for Radiance Field&lt;br&gt;
연구 홈페이지: &lt;a href=&#34;https://maincold2.github.io/c3dgs/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://maincold2.github.io/c3dgs/&lt;/a&gt;&lt;br&gt;
저자: 이주찬, 노다니엘, Xiangyu Sun, 고종환, 박은병&lt;/p&gt;
&lt;p&gt;자세한 내용은 성균관대 Research Stories에서 확인하실 수 있습니다: &lt;a href=&#34;https://www.skku.edu/skku/research/industry/researchStory_view.do?mode=view&amp;amp;articleNo=117212&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.skku.edu/skku/research/industry/researchStory_view.do?mode=view&amp;articleNo=117212&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab researchers Joochan Lee and Daniel Rho have had two papers selected as ICLR 2024 Spotlight and CVPR 2024 Highlight, respectively.&lt;/p&gt;
&lt;p&gt;This research was jointly supervised by Prof. Jong Hwan Ko and Prof. Eunbyung Park, presenting two methodologies for efficiently representing complex 3D scenes. The first is Coordinate-Aware Modulation, which fuses neural networks with grid-based representations, and the second is a compact 3D Gaussian representation technique.&lt;/p&gt;
&lt;p&gt;The Coordinate-Aware Modulation research fuses grid feature vectors through modulation at each layer of the neural network when representing 3D images or videos. While existing methods required large capacities, this research efficiently represented high-frequency signals using very small grids. When applied to various media data including images, videos, and 3D models, it demonstrated excellent signal reconstruction capabilities with small network sizes.&lt;/p&gt;
&lt;p&gt;The Compact 3D Gaussian Splatting research solved the storage capacity problem of representing scenes as 3D Gaussians. While recent 3D Gaussian methods enable fast rendering above 100 FPS, they required very large storage capacity. This research successfully reduced the number of Gaussians without performance degradation and achieved both high performance and efficient storage space through a new representation methodology. Evaluation on real datasets achieved over 25x storage reduction and rendering speed improvement without quality degradation.&lt;/p&gt;
&lt;p&gt;The first research was selected as Spotlight at ICLR 2024, a top machine learning conference, representing the top 6% of submissions. The second was selected as Highlight at CVPR 2024, a top computer vision conference, representing the top 3% of submissions. These efficient 3D representation technologies are expected to be practically applied in various fields such as NeRF, generative models, and metaverse. Building on this foundation, IRIS Lab plans to continue advancing research in 3D vision technologies.&lt;/p&gt;
&lt;p&gt;Paper 1: Coordinate-Aware Modulation for Neural Fields&lt;br&gt;
Research homepage: &lt;a href=&#34;https://maincold2.github.io/cam/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://maincold2.github.io/cam/&lt;/a&gt;&lt;br&gt;
Authors: Joochan Lee, Daniel Rho, Seungtae Nam, Jong Hwan Ko, Eunbyung Park&lt;/p&gt;
&lt;p&gt;Paper 2: Compact 3D Gaussian Representation for Radiance Field&lt;br&gt;
Research homepage: &lt;a href=&#34;https://maincold2.github.io/c3dgs/&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://maincold2.github.io/c3dgs/&lt;/a&gt;&lt;br&gt;
Authors: Joochan Lee, Daniel Rho, Xiangyu Sun, Jong Hwan Ko, Eunbyung Park&lt;/p&gt;
&lt;p&gt;For more details, please visit SKKU Research Stories: &lt;a href=&#34;https://www.skku.edu/skku/research/industry/researchStory_view.do?mode=view&amp;amp;articleNo=117212&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.skku.edu/skku/research/industry/researchStory_view.do?mode=view&amp;articleNo=117212&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C61] HandDiff: 3D Hand Pose Estimation with Diffusion on Image-Point Cloud</title>
      <link>https://iris-lab.skku.edu/publication/c61_cvpr_2024_hand/</link>
      <pubDate>Thu, 07 Mar 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c61_cvpr_2024_hand/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C60] Compact 3D Gaussian Representation for Radiance Field</title>
      <link>https://iris-lab.skku.edu/publication/c60_cvpr_2024/</link>
      <pubDate>Wed, 28 Feb 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c60_cvpr_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C59] ConvMapSim: Modeling and Simulating Convolutional Weight Mapping for PIM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c59_aicas_2024_pim/</link>
      <pubDate>Wed, 21 Feb 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c59_aicas_2024_pim/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C58] An Efficient Ventricular Arrhythmias Detection on Microcontrollers with Optimized 1D CNN</title>
      <link>https://iris-lab.skku.edu/publication/c58_aicas_2024_tinyml/</link>
      <pubDate>Tue, 20 Feb 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c58_aicas_2024_tinyml/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C57] Adaptive Image Downscaling for Rate-Accuracy-Latency Optimization of Task-Target Image Compression</title>
      <link>https://iris-lab.skku.edu/publication/c57_aicas_2024_adaptive/</link>
      <pubDate>Mon, 19 Feb 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c57_aicas_2024_adaptive/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C56] TraiNDSim: A Simulation Framework for Comprehensive Performance Evaluation of Neuromorphic Devices for On-Chip Training</title>
      <link>https://iris-lab.skku.edu/publication/c56_dac_2024/</link>
      <pubDate>Sun, 18 Feb 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c56_dac_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J28] KERNTROL: Kernel Shape Control Toward Ultimate Memory Utilization for In-Memory Convolutional Weight Mapping</title>
      <link>https://iris-lab.skku.edu/publication/j28_tcas1_2024/</link>
      <pubDate>Sat, 17 Feb 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j28_tcas1_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C55]KARS: Kernel-Grouping Aided Row-Skipping for SDK-based Weight Compression in PIM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c55_iscas_2024/</link>
      <pubDate>Wed, 17 Jan 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c55_iscas_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C54]Coordinate-Aware Modulation for Neural Fields</title>
      <link>https://iris-lab.skku.edu/publication/c54_iclr_2024/</link>
      <pubDate>Tue, 16 Jan 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c54_iclr_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J27] A DNN Partitioning Framework with Controlled Lossy Mechanisms for Edge-Cloud Collaborative Intelligence</title>
      <link>https://iris-lab.skku.edu/publication/j27_fgcs_2024/</link>
      <pubDate>Mon, 15 Jan 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j27_fgcs_2024/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Hong Kong Visiting Program (BK214)</title>
      <link>https://iris-lab.skku.edu/post/bk_hongkong/</link>
      <pubDate>Mon, 15 Jan 2024 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/bk_hongkong/</guid>
      <description>&lt;p&gt;IRIS 연구실의 전강은 박사후연구원 인솔 하에, 박주홍, 소재현, 이존이, 황찬욱 학생은 2023년 12월 31일부터 2024년 1월 15일까지 BK214 지원사업을 통해 홍콩에서 다양한 학술 교류 및 연구 활동을 수행하였습니다.&lt;/p&gt;
&lt;p&gt;우선, 참가자들은 HKU(Hong Kong University) 에서 열린 CPAL(Conference on Parsimony and Learning) 2024 학회에 참석하여, 머신러닝·신호처리·최적화 전반에 걸쳐 존재하는 간결성(parsimony)과 저차원 구조(low-dimensional structure)의 최신 연구 성과를 접하고, 세계 연구자들과 활발히 교류하는 기회를 가졌습니다.&lt;/p&gt;
&lt;p&gt;또한, 홍콩과학기술대학교(HKUST) Larry Li 교수 연구실과 연구 교류를 진행하며 열음향 불안정성(thermoacoustic instability) 문제를 학습하고, 1D CNN 기반 회귀 모델을 활용한 조기 탐지 실험을 수행하였습니다. 이 과정에서 박정진·허준 연구원과 협업하며 학제적 연구 경험을 쌓았습니다.&lt;/p&gt;
&lt;p&gt;마지막으로, 홍콩시립대학교(CityU) Arindam Basu 교수 연구실을 방문하여 양 연구실의 연구 방향을 공유하고, In-Memory Computing, 뉴로모픽 시스템, 뇌-기계 인터페이스(BMI) 분야 등에서의 협력 가능성을 논의하였습니다.&lt;/p&gt;
&lt;p&gt;본 BK214 교류 프로그램은 학회 참석·연구 교류·랩 방문 등 다층적 활동을 통해 IRIS Lab 학생들에게 국제 공동연구 역량 강화와 새로운 협력 기회 발굴의 소중한 경험을 제공하였습니다.&lt;/p&gt;
&lt;p&gt;Under the guidance of Dr. Kang Eun Jeon (Postdoctoral Researcher), IRIS Lab members Johnny Rhe, Jaehyeon So, Chanwook Hwang, and Juhong Park participated in the BK214-supported research exchange program in Hong Kong from December 31, 2023, to January 15, 2024.&lt;/p&gt;
&lt;p&gt;The program began with attendance at the Conference on Parsimony and Learning (CPAL) 2024, which explored parsimonious and low-dimensional structures in machine learning, signal processing, optimization, and related domains. The conference offered participants exposure to cutting-edge research and opportunities to engage with the global academic community.&lt;/p&gt;
&lt;p&gt;The team also engaged in a research exchange with Prof. Larry Li’s Lab at HKUST, where they studied thermoacoustic instability and implemented a 1D convolutional neural network regression model for early detection. Collaborating with Jungjin Park and Jun Hur, they combined expertise in physics and machine learning to advance their understanding.&lt;/p&gt;
&lt;p&gt;Finally, the group visited Prof. Arindam Basu’s Lab at CityU, exchanging research directions and exploring potential collaborations in neuromorphic systems and brain-machine interfaces (BMIs).&lt;/p&gt;
&lt;p&gt;Through this BK214 exchange program, IRIS Lab students gained invaluable experience in international research networking and cross-disciplinary collaboration, strengthening their capacity for future global research initiatives.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[J26] Highly Reliable 3D Channel Memory and Its Application in a Neuromorphic Sensory System for Finger Motion Tracking</title>
      <link>https://iris-lab.skku.edu/publication/j26_acs_nano_2023/</link>
      <pubDate>Wed, 20 Dec 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j26_acs_nano_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C53]Facto-CNN: Memory-Efficient CNN Training with Low-rank Tensor Factorization and Lossy Tensor Compression</title>
      <link>https://iris-lab.skku.edu/publication/c53_acml_2023/</link>
      <pubDate>Mon, 16 Oct 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c53_acml_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C52] Mip-Grid: Anti-aliased Grid Representations for Neural Radiance Fields</title>
      <link>https://iris-lab.skku.edu/publication/c52_nips_2023/</link>
      <pubDate>Sun, 15 Oct 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c52_nips_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C51] DCR: Decomposition-Aware Column Re-Mapping for Stuck-At-Fault Tolerance in ReRAM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c51_iccd_2023/</link>
      <pubDate>Fri, 28 Jul 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c51_iccd_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C50] FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos</title>
      <link>https://iris-lab.skku.edu/publication/c50_mm_2023/</link>
      <pubDate>Thu, 27 Jul 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c50_mm_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C49] Multi-Scale Bidirectional Recurrent Network with Hybrid Correlation for Point Cloud Based Scene Flow Estimation</title>
      <link>https://iris-lab.skku.edu/publication/c49_iccv_2023/</link>
      <pubDate>Wed, 26 Jul 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c49_iccv_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C48] HandR2N2: Iterative 3D Hand Pose Estimation Using a Residual Recurrent Neural Network</title>
      <link>https://iris-lab.skku.edu/publication/c48_iccv_2023/</link>
      <pubDate>Tue, 25 Jul 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c48_iccv_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C47] Kernel Shape Control for Row-Efficient Convolution on Processing-In-Memory Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c47_iccad_2023/</link>
      <pubDate>Fri, 21 Jul 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c47_iccad_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C46] An Energy Efficient Sorting Architecture with Cell-Gating for Top-K Sorting on FPGA</title>
      <link>https://iris-lab.skku.edu/publication/c46_mwscas_2023/</link>
      <pubDate>Sat, 10 Jun 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c46_mwscas_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C45] PAIRS: Pruning-AIded Row-Skipping for Convolutional Weight Mapping in Processing-In-Memory Architectures</title>
      <link>https://iris-lab.skku.edu/publication/c45_islped_2023/</link>
      <pubDate>Wed, 17 May 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c45_islped_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C44] Weight-Aware Activation Mapping for Energy-Efficient Convolution on PIM Arrays</title>
      <link>https://iris-lab.skku.edu/publication/c44_islped_2023/</link>
      <pubDate>Tue, 16 May 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c44_islped_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C43] Dynamic Inference Acceleration of 3D Point Cloud Deep Neural Networks Using Point Density and Entropy</title>
      <link>https://iris-lab.skku.edu/publication/c43_cvprw_2023/</link>
      <pubDate>Wed, 26 Apr 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c43_cvprw_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C42] Rate-Controllable and Target-Dependent JPEG-Based Image Compression Using Feature Modulation</title>
      <link>https://iris-lab.skku.edu/publication/c42_icme_2023/</link>
      <pubDate>Tue, 25 Apr 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c42_icme_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C41] Masked Wavelet Representation for Compact Neural Radiance Fields </title>
      <link>https://iris-lab.skku.edu/publication/c41_cvpr_2023/</link>
      <pubDate>Mon, 27 Feb 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c41_cvpr_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C40] Regression to Classification: Waveform Encoding for Neural Field-Based Audio Signal Representation</title>
      <link>https://iris-lab.skku.edu/publication/c40__icassp_2023/</link>
      <pubDate>Thu, 16 Feb 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c40__icassp_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J25] Data-Driven Design of Electrically Conductive Nanocomposite Materials: A Case Study of Acrylonitrile–Butadiene–Styrene/Carbon Nanotube Binary Composites</title>
      <link>https://iris-lab.skku.edu/publication/j25_ais_2023/</link>
      <pubDate>Wed, 18 Jan 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j25_ais_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J24] Room-Temperature-Processable Highly Reliable Resistive Switching Memory with Reconfigurability for Neuromorphic Computing and Ultrasonic Tissue Classification</title>
      <link>https://iris-lab.skku.edu/publication/j24_afm_2023/</link>
      <pubDate>Tue, 17 Jan 2023 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j24_afm_2023/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J23] An Overhead-Free Region-Based JPEG Framework for Task-Driven Image Compression</title>
      <link>https://iris-lab.skku.edu/publication/j23_prl_2022/</link>
      <pubDate>Thu, 17 Nov 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j23_prl_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C39] Element-wise Partial Product Quantization for Efficient Deep Learning Accelerators</title>
      <link>https://iris-lab.skku.edu/publication/c39_icce-asia_2022/</link>
      <pubDate>Thu, 27 Oct 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c39_icce-asia_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C38] An Efficient Systolic Array with Variable Data Precision and Dimension Support</title>
      <link>https://iris-lab.skku.edu/publication/c38_isocc_2022/</link>
      <pubDate>Fri, 21 Oct 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c38_isocc_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C37] Neural Residual Flow Fields for Efficient Video Representations</title>
      <link>https://iris-lab.skku.edu/publication/c37_accv_2022/</link>
      <pubDate>Fri, 16 Sep 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c37_accv_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J22] Micro-ultrasonic Assessment of Early Stage Clot Formation and Whole Blood Coagulation Using an All-Optical Ultrasound Transducer and an Adaptive Signal Processing Algorithm</title>
      <link>https://iris-lab.skku.edu/publication/j22_acs_sensors_2022/</link>
      <pubDate>Thu, 08 Sep 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j22_acs_sensors_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J21] A Reconfigurable Neural Architecture for Edge–Cloud Collaborative Real-Time Object Detection</title>
      <link>https://iris-lab.skku.edu/publication/j21_iot_2022/</link>
      <pubDate>Mon, 05 Sep 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j21_iot_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C36] Bi-PointFlowNet: Bidirectional Learning for Point Cloud Based Scene Flow Estimation</title>
      <link>https://iris-lab.skku.edu/publication/c36_eccv_2022/</link>
      <pubDate>Mon, 11 Jul 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c36_eccv_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C35] Streamable Neural Fields</title>
      <link>https://iris-lab.skku.edu/publication/c35_eccv_2022/</link>
      <pubDate>Sun, 10 Jul 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c35_eccv_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J20] Scalable Color Quantization for Task-Centric Image Compression</title>
      <link>https://iris-lab.skku.edu/publication/j20_tomm_2022/</link>
      <pubDate>Mon, 20 Jun 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j20_tomm_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J19] Epidermal Piezoresistive Structure with Deep Learning-Assisted Data Translation</title>
      <link>https://iris-lab.skku.edu/publication/j19_npj_2022/</link>
      <pubDate>Wed, 15 Jun 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j19_npj_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C34] NAS-VAD: Neural Architecture Search for Voice Activity Detection</title>
      <link>https://iris-lab.skku.edu/publication/c34_interspeech-2022/</link>
      <pubDate>Wed, 01 Jun 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c34_interspeech-2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J18] Multi-Prediction Compression: An Efficient and Scalable Memory Compression Framework for GP-GPU</title>
      <link>https://iris-lab.skku.edu/publication/j18_cal_2022/</link>
      <pubDate>Sat, 21 May 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j18_cal_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J17] VWC-SDK: Convolutional Weight Mapping Using Shifted and Duplicated Kernel with Variable Windows and Channels</title>
      <link>https://iris-lab.skku.edu/publication/j17_jetcas_2022/</link>
      <pubDate>Sun, 01 May 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j17_jetcas_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C33] ADA-VAD: Unpaired Adversarial Domain Adaptation for Noise-Robust Voice Activity Detection</title>
      <link>https://iris-lab.skku.edu/publication/c33_icassp_2022/</link>
      <pubDate>Tue, 01 Feb 2022 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c33_icassp_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C32] VW-SDK: Efficient Convolutional Weight Mapping Using Variable Windows for Processing-In-Memory Architectures</title>
      <link>https://iris-lab.skku.edu/publication/c32_date_2022/</link>
      <pubDate>Thu, 11 Nov 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c32_date_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J16] ORVAE: One-Class Residual Variational Autoencoder for Voice Activity Detection in Noisy Environment</title>
      <link>https://iris-lab.skku.edu/publication/j16_npl_2022/</link>
      <pubDate>Wed, 10 Nov 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j16_npl_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J15] Adaptive Weight-bit Inversion for State Error Reduction for Robust and Efficient Deep Neural Network Inference Using MLC NAND Flash</title>
      <link>https://iris-lab.skku.edu/publication/j15_jsa_2022/</link>
      <pubDate>Mon, 01 Nov 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j15_jsa_2022/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C31] HandFoldingNet: A 3D Hand Pose Estimation Network Using Multiscale-Feature Guided Folding of a 2D Hand Skeleton</title>
      <link>https://iris-lab.skku.edu/publication/c31_iccv_2021/</link>
      <pubDate>Fri, 01 Oct 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c31_iccv_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J14] Binary-Classifiers-Enabled Filters for Semi-Supervised Learning</title>
      <link>https://iris-lab.skku.edu/publication/j14_access_2021/</link>
      <pubDate>Fri, 01 Oct 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j14_access_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C30] A Splittable DNN-Based Object Detector for Edge-Cloud Collaborative Real-Time Video Inference</title>
      <link>https://iris-lab.skku.edu/publication/c30_avss_2021/</link>
      <pubDate>Fri, 17 Sep 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c30_avss_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C29] A Charge-Domain Computation-in-Memory Macro with Versatile All-Around-Wire-Capacitor for Variable-Precision Computation and Array-Embedded DA/AD Conversions</title>
      <link>https://iris-lab.skku.edu/publication/c29_esscirc_2021/</link>
      <pubDate>Wed, 15 Sep 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c29_esscirc_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J13] Target-Dependent Scalable Image Compression Using a Reconfigurable Recurrent Neural Network</title>
      <link>https://iris-lab.skku.edu/publication/j13_access_2021/</link>
      <pubDate>Wed, 01 Sep 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j13_access_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C28] A Sub-Milliwatt and Sub-Millisecond 3-D GazeEstimator for Ultra Low-Power AR Applications</title>
      <link>https://iris-lab.skku.edu/publication/c28_eyeware_2021/</link>
      <pubDate>Sat, 10 Jul 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c28_eyeware_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[B01] Ultra-low Power/IoT Architectures</title>
      <link>https://iris-lab.skku.edu/publication/b01_iot_2021/</link>
      <pubDate>Wed, 26 May 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/b01_iot_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J12] Robust Detection of Small and Dense Objects in Images from Autonomous Aerial Vehicles</title>
      <link>https://iris-lab.skku.edu/publication/j12_iet_el_2021/</link>
      <pubDate>Sat, 15 May 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j12_iet_el_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J11] A Charge-Domain Scalable-Weight In-Memory Computing Macro with Dual-SRAM Architecture for Precision-Scalable DNN Accelerators</title>
      <link>https://iris-lab.skku.edu/publication/j11_tcas1_2021/</link>
      <pubDate>Sat, 01 May 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j11_tcas1_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C27] Application of Adversarial Domain Adaptation to Voice Activity Detection</title>
      <link>https://iris-lab.skku.edu/publication/c27_intellisys_2021/</link>
      <pubDate>Thu, 15 Apr 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c27_intellisys_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J10] Segmentation of Points in the Future: Joint Segmentation and Prediction of a Point Cloud</title>
      <link>https://iris-lab.skku.edu/publication/j10_access_2021/</link>
      <pubDate>Sun, 14 Mar 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j10_access_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C26] WISER: Deep Neural Network Weight-bit Inversion for State Error Reduction in MLC NAND Flash</title>
      <link>https://iris-lab.skku.edu/publication/c26_date_2021/</link>
      <pubDate>Mon, 15 Feb 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c26_date_2021/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Wins 2020 AI Grand Challenge and MSIT Minister&#39;s Award</title>
      <link>https://iris-lab.skku.edu/post/ai_grand_challenge_2020/</link>
      <pubDate>Mon, 04 Jan 2021 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/ai_grand_challenge_2020/</guid>
      <description>&lt;p&gt;IRIS 연구실이 과학기술정보통신부 주관 &amp;lsquo;2020 인공지능 R&amp;amp;D 그랜드 챌린지&amp;rsquo; 2단계 대회에서 우승을 차지했습니다. 김태수, 노다니엘, 이승진 연구원으로 구성된 팀이 음향인지 트랙에서 1위를 기록하며 과기정통부 장관상과 함께 후속연구비 7억 원을 지원받게 되었습니다.&lt;/p&gt;
&lt;p&gt;AI R&amp;amp;D 그랜드 챌린지는 제시된 도전과제를 해결하는 알고리즘을 실제로 개발하여 성능으로 경쟁하는 대회입니다. 2019년 1단계 대회를 시작으로 2022년까지 총 4단계로 진행되며, 각 단계마다 과제 난이도가 높아집니다.&lt;/p&gt;
&lt;p&gt;이번 2단계 대회의 음향인지 트랙 과제는 드론의 소음 속에서 사람의 구조요청 소리를 감지하고 발원 방향을 추정하는 것이었습니다. 1단계보다 더 복잡한 소음 환경과 다양한 거리, 각도 조건이 추가되어 난이도가 크게 높아졌습니다. IRIS 연구실은 높은 점수로 1위를 차지하며 기술력을 입증했습니다.&lt;/p&gt;
&lt;p&gt;연구실은 2019년 1단계 대회에서 3위를 차지한 뒤, 그 경험을 바탕으로 알고리즘을 개선하고 더욱 견고한 시스템을 구축해 왔습니다. 1년여의 연구 끝에 2단계 대회에서 우승을 이루어낸 것은, 지속적인 기술 발전과 연구팀의 노력이 만들어낸 성과였습니다.&lt;/p&gt;
&lt;p&gt;드론 소음 환경에서의 음향 인식 기술은 재난 현장에서 생존자를 찾거나, 위급 상황을 신속하게 파악하는 데 활용될 수 있습니다. 이번 대회를 통해 실전에 가까운 조건에서 기술을 검증하고 발전시킨 경험은, 향후 실용적인 오디오 AI 시스템 개발로 이어질 수 있을 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;자세한 내용은 정보통신대학 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=110273&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=110273&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab won first place in Stage 2 of the &amp;lsquo;2020 AI R&amp;amp;D Grand Challenge&amp;rsquo; hosted by the Ministry of Science and ICT. The team comprising researchers Taesoo Kim, Daniel Rho, and Seungjin Lee took first place in the auditory cognition track, earning the MSIT Minister&amp;rsquo;s Award along with 700 million KRW in follow-up research funding.&lt;/p&gt;
&lt;p&gt;The AI R&amp;amp;D Grand Challenge is a competition where teams develop algorithms to solve presented challenges and compete based on actual performance. Running from the 2019 Stage 1 through 2022 across four stages, the difficulty increases with each stage.&lt;/p&gt;
&lt;p&gt;The Stage 2 auditory cognition track task required detecting human distress calls in drone noise and estimating their direction of origin. With more complex noise environments and various distance and angle conditions added compared to Stage 1, the difficulty increased significantly. IRIS Lab demonstrated its technical capabilities by achieving first place with a high score.&lt;/p&gt;
&lt;p&gt;After placing 3rd in the 2019 Stage 1 competition, the lab improved its algorithms and built more robust systems based on that experience. Achieving victory in Stage 2 after a year of research represents an outcome created by continuous technological advancement and the research team&amp;rsquo;s dedication.&lt;/p&gt;
&lt;p&gt;Audio recognition technology in drone noise environments can be utilized to locate survivors at disaster sites or rapidly identify emergency situations. The experience of validating and advancing the technology under near-real-world conditions through this competition is expected to lead to the development of practical audio AI systems in the future.&lt;/p&gt;
&lt;p&gt;For more details, please visit the official news from the College of Information &amp;amp; Communication Engineering: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=110273&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=110273&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Neuromorphic/In-Memory Computing</title>
      <link>https://iris-lab.skku.edu/project/ml_mem/</link>
      <pubDate>Tue, 22 Dec 2020 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/ml_mem/</guid>
      <description>&lt;p&gt;Our research focuses on memory-based deep learning techniques, including SRAM/ReRAM-based deep learning processing-in-memory (PIM) and Flash memory based robust deep learning inference techniques.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Machine Learning for 3D Data Processing</title>
      <link>https://iris-lab.skku.edu/project/mm_3d/</link>
      <pubDate>Sun, 22 Nov 2020 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/mm_3d/</guid>
      <description>&lt;p&gt;As three-dimensional (3D) vision data can provide abundant spatial information, it is being widely used in many
areas, including autonomous driving and mobile robots. We aim to develop efficient and accurate 3D data processing techniques for classification, segmentation, pose estimation, etc.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>IRIS Lab Wins 1st Place at ECCV 2020 VisDrone Challenge</title>
      <link>https://iris-lab.skku.edu/post/eccv_visdrone_challenge_2020/</link>
      <pubDate>Wed, 14 Oct 2020 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/eccv_visdrone_challenge_2020/</guid>
      <description>&lt;p&gt;IRIS 연구실이 컴퓨터 비전 분야 학회인 ECCV가 주관하는 &amp;lsquo;VisDrone Challenge 2020&amp;rsquo; 객체 탐지 부문에서 1위를 차지했습니다. 이주찬 연구원과 유정엽 연구원이 한국항공우주연구원과 함께 팀 &amp;lsquo;DroneEye 2020&amp;rsquo;으로 참가한 이번 대회에서, 전 세계 36개 팀을 제치고 1위를 기록했습니다.&lt;/p&gt;
&lt;p&gt;VisDrone Challenge는 ICCV와 ECCV가 2018년부터 매년 개최하는 국제 대회로, 드론이 촬영한 영상에서 사람이나 자동차 등 객체를 인공지능으로 탐지하는 기술을 겨룹니다. 드론 영상은 비행 고도에 따라 객체의 크기가 크게 변하고, 각도와 조명 조건도 불규칙하게 바뀌기 때문에 일반 영상 인식보다 어려운 과제입니다.&lt;/p&gt;
&lt;p&gt;이번 대회에는 미국 노스캐롤라이나대학, 독일 프라운호퍼 연구소, 중국과학기술대 등의 연구기관들이 참가했습니다. IRIS 연구실은 드론의 비행 고도에 따라 변화하는 지상 객체들을 다중으로 인식해 내는 기술을 개발했으며, 다양한 비행 환경과 조건에서도 안정적으로 작동하는 알고리즘을 제시했습니다.&lt;/p&gt;
&lt;p&gt;연구실은 2019년 AI R&amp;amp;D 그랜드챌린지에서 오디오 인식 트랙 3위, 2020년 동 대회에서 이미지 인식 트랙 입상에 이어, 이번 VisDrone Challenge에서 비디오 인식 분야 1위를 차지하며 오디오, 이미지, 비디오를 아우르는 멀티미디어 AI 기술을 축적해 왔습니다.&lt;/p&gt;
&lt;p&gt;이번 우승 소식은 연합뉴스, 매일경제, MBN, 헤럴드경제, 아시아경제, 동아사이언스 등 여러 언론에서 보도되었습니다. 드론 기반 객체 인식 기술은 재난 감시, 교통 모니터링, 스마트 시티 관리 등에 응용될 수 있는 분야로, 이번 대회를 통해 얻은 경험이 향후 실용적인 드론 비전 시스템 연구와 개발에 기여할 수 있기를 기대합니다.&lt;/p&gt;
&lt;p&gt;관련 기사:&lt;br&gt;
연합뉴스: &lt;a href=&#34;https://www.yna.co.kr/view/AKR20201006084700063?input=1195m&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.yna.co.kr/view/AKR20201006084700063?input=1195m&lt;/a&gt;&lt;br&gt;
매일경제: &lt;a href=&#34;http://mbnmoney.mbn.co.kr/news/view?news_no=MM1004131427&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;http://mbnmoney.mbn.co.kr/news/view?news_no=MM1004131427&lt;/a&gt;&lt;br&gt;
MBN: &lt;a href=&#34;https://www.mbn.co.kr/news/economy/4301197&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.mbn.co.kr/news/economy/4301197&lt;/a&gt;&lt;br&gt;
헤럴드경제: &lt;a href=&#34;http://news.heraldcorp.com/view.php?ud=20201006000942&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;http://news.heraldcorp.com/view.php?ud=20201006000942&lt;/a&gt;&lt;br&gt;
아시아경제: &lt;a href=&#34;https://view.asiae.co.kr/article/2020100613200587789&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://view.asiae.co.kr/article/2020100613200587789&lt;/a&gt;&lt;br&gt;
동아사이언스: &lt;a href=&#34;http://dongascience.donga.com/news/view/40368&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;http://dongascience.donga.com/news/view/40368&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;자세한 내용은 정보통신대학 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=102152&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=102152&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab won first place in the object detection track at &amp;lsquo;VisDrone Challenge 2020&amp;rsquo; hosted by ECCV, a leading computer vision conference. Researchers Joochan Lee and Jungyeop Yoo participated as team &amp;lsquo;DroneEye 2020&amp;rsquo; with the Korea Aerospace Research Institute, achieving first place among 36 teams worldwide.&lt;/p&gt;
&lt;p&gt;The VisDrone Challenge, held annually since 2018 by ICCV and ECCV, is an international competition focused on AI-based object detection in drone-captured footage. Drone video presents greater challenges than conventional video recognition due to significant variations in object size depending on flight altitude, along with irregular changes in viewing angles and lighting conditions.&lt;/p&gt;
&lt;p&gt;Participants included research institutions such as the University of North Carolina, Fraunhofer Institute in Germany, and University of Science and Technology of China. IRIS Lab developed technology to detect multiple ground objects that vary with drone flight altitude, presenting an algorithm that operates stably across diverse flight environments and conditions.&lt;/p&gt;
&lt;p&gt;The lab has accumulated multimodal AI technologies spanning audio, image, and video: placing 3rd in the audio recognition track at the 2019 AI R&amp;amp;D Grand Challenge, winning recognition in the image recognition track at the 2020 competition, and now taking 1st place in video recognition at the VisDrone Challenge.&lt;/p&gt;
&lt;p&gt;The victory was covered by various media outlets including Yonhap News, Maeil Business, MBN, Herald Economy, Asia Economy, and Donga Science. Drone-based object recognition technology has applications in disaster monitoring, traffic surveillance, and smart city management. We hope the experience gained through this competition will contribute to future research and development of practical drone vision systems.&lt;/p&gt;
&lt;p&gt;Related articles:&lt;br&gt;
Yonhap News: &lt;a href=&#34;https://www.yna.co.kr/view/AKR20201006084700063?input=1195m&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.yna.co.kr/view/AKR20201006084700063?input=1195m&lt;/a&gt;&lt;br&gt;
Maeil Business: &lt;a href=&#34;http://mbnmoney.mbn.co.kr/news/view?news_no=MM1004131427&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;http://mbnmoney.mbn.co.kr/news/view?news_no=MM1004131427&lt;/a&gt;&lt;br&gt;
MBN: &lt;a href=&#34;https://www.mbn.co.kr/news/economy/4301197&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://www.mbn.co.kr/news/economy/4301197&lt;/a&gt;&lt;br&gt;
Herald Economy: &lt;a href=&#34;http://news.heraldcorp.com/view.php?ud=20201006000942&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;http://news.heraldcorp.com/view.php?ud=20201006000942&lt;/a&gt;&lt;br&gt;
Asia Economy: &lt;a href=&#34;https://view.asiae.co.kr/article/2020100613200587789&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://view.asiae.co.kr/article/2020100613200587789&lt;/a&gt;&lt;br&gt;
Donga Science: &lt;a href=&#34;http://dongascience.donga.com/news/view/40368&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;http://dongascience.donga.com/news/view/40368&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;For more details, please visit the official news from the College of Information &amp;amp; Communication Engineering: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=102152&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=102152&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C25] VisDrone-DET2020: The Vision Meets Drone Object Detection in Image Challenge Results</title>
      <link>https://iris-lab.skku.edu/publication/c25_eccv_2020/</link>
      <pubDate>Tue, 01 Sep 2020 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c25_eccv_2020/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Sensor Systems with Integrated Deep Learning</title>
      <link>https://iris-lab.skku.edu/project/hw_mlsensor/</link>
      <pubDate>Sun, 22 Dec 2019 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/hw_mlsensor/</guid>
      <description>&lt;p&gt;To maximize the performance and efficiency of deep learning based data processing, we are currently exploring sensor platform design optimized for deep learning by leveraging the interactions between a sensor platform and a DNN.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[J8] CAMEL: An Adaptive Camera with Embedded Machine Learning Based Sensor Parameter Control</title>
      <link>https://iris-lab.skku.edu/publication/j8_jetcas_2019/</link>
      <pubDate>Wed, 14 Aug 2019 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j8_jetcas_2019/</guid>
      <description></description>
    </item>
    
    <item>
      <title>IRIS Lab Wins 3rd Place at 2019 AI Grand Challenge</title>
      <link>https://iris-lab.skku.edu/post/ai_grand_challenge_2019/</link>
      <pubDate>Fri, 12 Jul 2019 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/post/ai_grand_challenge_2019/</guid>
      <description>&lt;p&gt;IRIS 연구실이 과학기술정보통신부 주관 &amp;lsquo;2019 인공지능 R&amp;amp;D 그랜드 챌린지&amp;rsquo;에서 3위에 입상했습니다. 이번 입상으로 상금 100만 원과 함께 향후 2년간 총 8억 원의 후속연구비를 지원받게 되었습니다.&lt;/p&gt;
&lt;p&gt;이번 대회는 일반적인 연구 제안서 심사 방식이 아닌, 실제 알고리즘을 개발해 성능으로 겨루는 경진대회 형식으로 진행되었습니다. 올해 첫 대회에는 전국의 대학, 연구소, 기업에서 121개 팀, 617명이 참가했으며, 각 트랙별로 상위 3개 팀만이 후속 연구 지원을 받게 됩니다.&lt;/p&gt;
&lt;p&gt;연구실이 도전한 청각인지 트랙의 과제는 소음이 가득한 환경에서 사람의 구조요청 소리를 찾아내고, 그 소리가 어디서 들려오는지 방향까지 추정하는 것이었습니다. 재난 현장이나 복잡한 도심 환경을 가정한 이 과제는 실제 구조 상황에서 활용될 수 있는 기술을 요구했습니다. 33개 팀이 경쟁한 이 트랙에서 IRIS 연구실은 3위를 차지했습니다.&lt;/p&gt;
&lt;p&gt;특히 이번 대회 참가팀은 대부분 학부생들로 구성되어 있었습니다. 전문 연구소나 기업팀들과 어깨를 나란히 하며 경쟁한 결과였기에 더욱 의미 있는 성과였습니다. 연구실은 평소 학부생들에게 실전 연구 경험을 쌓을 수 있는 기회를 제공하고 있으며, 이번 입상은 그러한 교육 철학의 결과이기도 합니다.&lt;/p&gt;
&lt;p&gt;이번 챌린지를 통해 소음 환경에서의 음성 인식과 방향 추정 기술을 실제 문제에 적용하며 얻은 경험은, 향후 연구실의 오디오 AI 연구와 실용화 과정에서 귀중한 자산이 될 것으로 기대됩니다.&lt;/p&gt;
&lt;p&gt;관련 기사:&lt;br&gt;
&lt;a href=&#34;https://news.naver.com/main/read.nhn?mode=LSD&amp;amp;mid=sec&amp;amp;sid1=105&amp;amp;oid=029&amp;amp;aid=0002544735&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://news.naver.com/main/read.nhn?mode=LSD&amp;mid=sec&amp;sid1=105&amp;oid=029&amp;aid=0002544735&lt;/a&gt;&lt;br&gt;
&lt;a href=&#34;https://news.naver.com/main/read.nhn?mode=LSD&amp;amp;mid=sec&amp;amp;sid1=102&amp;amp;oid=002&amp;amp;aid=0002099610&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://news.naver.com/main/read.nhn?mode=LSD&amp;mid=sec&amp;sid1=102&amp;oid=002&amp;aid=0002099610&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;자세한 내용은 정보통신대학 공식 소식에서 확인하실 수 있습니다: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=72287&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=72287&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;IRIS Lab won 3rd place at the &amp;lsquo;2019 AI R&amp;amp;D Grand Challenge&amp;rsquo; hosted by the Ministry of Science and ICT. The award includes a prize of 1 million KRW and follow-up research funding totaling 800 million KRW over the next two years.&lt;/p&gt;
&lt;p&gt;Unlike typical research proposal reviews, this competition was conducted as a challenge where teams compete based on actual algorithm performance. This year&amp;rsquo;s inaugural event saw 121 teams and 617 participants from universities, research institutes, and companies nationwide, with only the top three teams in each track receiving follow-up research support.&lt;/p&gt;
&lt;p&gt;The auditory cognition track challenged teams to detect human distress calls in noisy environments and estimate their direction. Simulating disaster sites or complex urban environments, the task required technology applicable to real rescue situations. IRIS Lab placed 3rd among 33 competing teams in this track.&lt;/p&gt;
&lt;p&gt;The team was primarily composed of undergraduate students, making the achievement particularly meaningful as they competed alongside professional research institutes and corporate teams. The lab regularly provides undergraduate students with hands-on research opportunities, and this award reflects that educational philosophy.&lt;/p&gt;
&lt;p&gt;The experience gained through this challenge in applying voice recognition and direction estimation technologies to real-world problems will serve as a valuable asset for the lab&amp;rsquo;s future audio AI research and practical applications.&lt;/p&gt;
&lt;p&gt;Related articles:&lt;br&gt;
&lt;a href=&#34;https://news.naver.com/main/read.nhn?mode=LSD&amp;amp;mid=sec&amp;amp;sid1=105&amp;amp;oid=029&amp;amp;aid=0002099610&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://news.naver.com/main/read.nhn?mode=LSD&amp;mid=sec&amp;sid1=105&amp;oid=029&amp;aid=0002099610&lt;/a&gt;&lt;br&gt;
&lt;a href=&#34;https://news.naver.com/main/read.nhn?mode=LSD&amp;amp;mid=sec&amp;amp;sid1=102&amp;amp;oid=002&amp;amp;aid=0002099610&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://news.naver.com/main/read.nhn?mode=LSD&amp;mid=sec&amp;sid1=102&amp;oid=002&amp;aid=0002099610&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;For more details, please visit the official news from the College of Information &amp;amp; Communication Engineering: &lt;a href=&#34;https://ice.skku.edu/ice/news.do?mode=view&amp;amp;articleNo=72287&#34; target=&#34;_blank&#34; rel=&#34;noopener&#34;&gt;https://ice.skku.edu/ice/news.do?mode=view&amp;articleNo=72287&lt;/a&gt;&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C24] A Camera with Brain–Embedding Machine Learning in 3D Sensors</title>
      <link>https://iris-lab.skku.edu/publication/c24_date_2019/</link>
      <pubDate>Mon, 25 Mar 2019 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c24_date_2019/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C23] Mixture of Pre-processing Experts Model for Noise Robust Deep Learning on Resource Constrained Platforms</title>
      <link>https://iris-lab.skku.edu/publication/c23_ijcnn_2019/</link>
      <pubDate>Fri, 01 Mar 2019 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c23_ijcnn_2019/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Deep Learning Model Compression and Acceleration</title>
      <link>https://iris-lab.skku.edu/project/ml_acceleration/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/ml_acceleration/</guid>
      <description>&lt;p&gt;Deep neural networks (DNNs) are widely adopted at the IoT edge devices to enable more intelligence.
However, the biggest challenge is their storage demand and computational complexity. As
DNNs contain a large number of synaptic weights, the memory demand is a key challenge
for application of DNNs, especially for memory-constrained platforms such as mobile systems.
Another bottleneck of DNNs is the computation demand, mostly due to their large
number of multiplications in convolution layers. Therefore, reducing the storage and computation
demand of neural networks is critical, specifically, to support in-field and on-chip
training and inference.
To enable deep learning based intelligent multimedia processing at the IoT edge devices
with limited hardware resource, the research targets for neural network design with lower
storage and computation demand.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Digital Design Techniques</title>
      <link>https://iris-lab.skku.edu/project/hw_digital/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/hw_digital/</guid>
      <description>&lt;p&gt;Tolerating timing error due to power supply
noise (PSN) in digital circuits can be done with adding voltage
margins. We designed guidelines to avoid overdesign
due to PSN especially for the low-cost IoT devices.
Specifically, we presented an accurate time-domain behavioral
model of timing slack variation due to PSN accounting for the
clock-data compensation.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Edge-Host Collaborative Deep Learning</title>
      <link>https://iris-lab.skku.edu/project/ml_collaboration/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/ml_collaboration/</guid>
      <description>&lt;p&gt;Based on the trade-off study between energy, accuracy, and throughput of an edge platform with an embedded DNN, we are currently exploring collaborative DNN inference between the edge and the host to optimally utilize the available bandwidth and energy.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Low-power Crypto Engines</title>
      <link>https://iris-lab.skku.edu/project/hw_security/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/hw_security/</guid>
      <description>&lt;p&gt;The physical security is a key challenge for the resource-constrained edge platforms.
A key challenge is to enable secure as well as ultra-low-power hardware.
The research seeks to understand the interactions between low-power and security
in edge devices, and explore innovations to enhance security at minimal power cost.
Specifically, the research investigated the interactions between design of
area-/power- efficient crypto engines and their side-channel resistance.
The goal is to enable side-channel secure encryption in resource-constrained platforms such as sensors.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Low-Power Image Sensor System</title>
      <link>https://iris-lab.skku.edu/project/hw_sensor/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/hw_sensor/</guid>
      <description>&lt;p&gt;The objective of this work is to design a self-powered, environment-adaptive sensor node that maintains a target Quality-of-Service (QoS) in a time-varying environment. A wireless image sensor node will be designed that incorporates a CMOS imager, digital signal processing unit, and RF transreceiver and is powered using energy harvested from the environment. The circuit innovations for the individual components will be coupled with system design and on-line real-time control principles to approach this highly challenging goal of developing a self-powered image sensor node. The self-powered sensor node and reliable energy-efficient image transmission principles created in this work will allow deployment of image sensors and communication networks to cyber-physical systems in various military as well as civilian applications.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Machine Learning for Audio/Speech Processing</title>
      <link>https://iris-lab.skku.edu/project/mm_audio/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/mm_audio/</guid>
      <description>&lt;p&gt;Our research focuses on front-end audio processing techniques for speech processing, such as voice activity detection (VAD), noise suppression, direction of arrival (DoA) estimation, and classification.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Machine Learning for Image/Video Processing</title>
      <link>https://iris-lab.skku.edu/project/mm_image/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/mm_image/</guid>
      <description>&lt;p&gt;Machine learning and deep learning have made rapid progress in many computer vision applications over a short period. We focus on diverse techniques for image and video processing powered by machine/deep learning.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>Variation-Robust Deep Learning</title>
      <link>https://iris-lab.skku.edu/project/ml_robust/</link>
      <pubDate>Sat, 22 Dec 2018 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/ml_robust/</guid>
      <description>&lt;p&gt;One of the challenges of deploying deep neural networks in sensor platforms is the variations
in the input images; structural noise (adversarial images), inherent random noise (image perturbation), input scene
variability, and weight value errors. To enhance the robustness of deep learning inference to the variability, we are designing variation-robust deep learning techniques.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C22] Edge-Host Partitioning of Deep Neural Networks with Feature Space Encoding for Resource-Constrained Internet-of-Things Platforms</title>
      <link>https://iris-lab.skku.edu/publication/c22_avss_2018/</link>
      <pubDate>Thu, 22 Nov 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c22_avss_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C21] An Energy Harvesting Image Sensor SOC with 5.8μW/mm2 Power Generation and 0.77 frames/second Self-Powered Frame Rate</title>
      <link>https://iris-lab.skku.edu/publication/c21_s3s_2018/</link>
      <pubDate>Tue, 16 Oct 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c21_s3s_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J7] Energy Efficient and Side-Channel Secure Cryptographic Hardware for IoT-edge Nodes</title>
      <link>https://iris-lab.skku.edu/publication/j7_iot_2018/</link>
      <pubDate>Tue, 31 Jul 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j7_iot_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C20] Edge-cloud collaborative processing for intelligent internet of things: a case study on smart surveillance</title>
      <link>https://iris-lab.skku.edu/publication/c20_dac_2018/</link>
      <pubDate>Sun, 24 Jun 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c20_dac_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J6] 3-D Stacked Image Sensor With Deep Neural Network Computation</title>
      <link>https://iris-lab.skku.edu/publication/j6_sensors_2018/</link>
      <pubDate>Tue, 15 May 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j6_sensors_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J5] An Energy-Quality Scalable Wireless Image Sensor Node for Object-Based Video Surveillance</title>
      <link>https://iris-lab.skku.edu/publication/j5_jetcas_2018/</link>
      <pubDate>Wed, 02 May 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j5_jetcas_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C19] Cascade Adversarial Machine Learning Regularized with a Unified Embedding</title>
      <link>https://iris-lab.skku.edu/publication/c19_iclr_2018/</link>
      <pubDate>Mon, 30 Apr 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c19_iclr_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C18] An Unsupervised Anomalous Event Detection Framework with Class Aware Source Separation</title>
      <link>https://iris-lab.skku.edu/publication/c18_icassp_2018_anomaly/</link>
      <pubDate>Wed, 18 Apr 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c18_icassp_2018_anomaly/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C17] Precision Scaling of Neural Networks for Efficient Voice Activity Detection</title>
      <link>https://iris-lab.skku.edu/publication/c17_icassp_2018_vad/</link>
      <pubDate>Tue, 17 Apr 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c17_icassp_2018_vad/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C16] A Unified Embedding for Image Classification and Pixel-Level Regularization</title>
      <link>https://iris-lab.skku.edu/publication/c16_icassp_2018_embedding/</link>
      <pubDate>Mon, 16 Apr 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c16_icassp_2018_embedding/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C15] The CAMEL approach to stacked sensor smart cameras</title>
      <link>https://iris-lab.skku.edu/publication/c15_date_2018/</link>
      <pubDate>Thu, 22 Mar 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c15_date_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J4] Design and Analysis of a Neural Network Inference Engine based on Adaptive Weight Compression</title>
      <link>https://iris-lab.skku.edu/publication/j4_tcad_2018/</link>
      <pubDate>Fri, 02 Feb 2018 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j4_tcad_2018/</guid>
      <description></description>
    </item>
    
    <item>
      <title>Low-Power Image/Video Processing</title>
      <link>https://iris-lab.skku.edu/project/mm_lowpower/</link>
      <pubDate>Fri, 22 Dec 2017 22:12:44 +0900</pubDate>
      <guid>https://iris-lab.skku.edu/project/mm_lowpower/</guid>
      <description>&lt;p&gt;A critical goal in the image sensor node design is to deliver high-quality visual information under stringent energy and bandwidth constraints. This goal becomes more challenging under dynamic conditions such as environmental noise and variations in a wireless channel condition. We tackle this challenge by designing low-power energy-quality scalable image processing techniques.&lt;/p&gt;
</description>
    </item>
    
    <item>
      <title>[C14] Precision Scaling of Neural Networks for Efficient Audio Processing</title>
      <link>https://iris-lab.skku.edu/publication/c14_nips_2017_speech/</link>
      <pubDate>Thu, 07 Dec 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c14_nips_2017_speech/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C13] Cascade Adversarial Machine Learning Regularized with a Unified Embedding</title>
      <link>https://iris-lab.skku.edu/publication/c13_nips_2017_adversarial/</link>
      <pubDate>Wed, 06 Dec 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c13_nips_2017_adversarial/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C12] Energy-Efficient Neural Image Processing for Internet-of-Things Edge Devices</title>
      <link>https://iris-lab.skku.edu/publication/c12_mwscas_2017/</link>
      <pubDate>Mon, 07 Aug 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c12_mwscas_2017/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C11] Design of an Energy-Efficient Accelerator for Training of Convolutional Neural Networks using Frequency-Domain Computation</title>
      <link>https://iris-lab.skku.edu/publication/c11_dac_2017/</link>
      <pubDate>Mon, 26 Jun 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c11_dac_2017/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J3] A Single-Chip Image Sensor Node With Energy Harvesting From a CMOS Pixel Array</title>
      <link>https://iris-lab.skku.edu/publication/j3_tcas_2017_sensor/</link>
      <pubDate>Sun, 18 Jun 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j3_tcas_2017_sensor/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J2] Clock Data Compensation Aware Digital Circuits Design for Voltage Margin Reduction</title>
      <link>https://iris-lab.skku.edu/publication/j2_tcas_2017_clock/</link>
      <pubDate>Sat, 17 Jun 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j2_tcas_2017_clock/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C10] On-Chip Training of Recurrent Neural Networks with Limited Numerical Precision</title>
      <link>https://iris-lab.skku.edu/publication/c10_ijcnn_2017/</link>
      <pubDate>Wed, 24 May 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c10_ijcnn_2017/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C9] Adaptive Weight Compression for Memory-Efficient Neural Networks</title>
      <link>https://iris-lab.skku.edu/publication/c9_date_2017_dnn/</link>
      <pubDate>Sat, 25 Mar 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c9_date_2017_dnn/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C8] Clock Data Compensation Aware Clock Tree Synthesis in Digital Circuits with Adaptive Clock Generation</title>
      <link>https://iris-lab.skku.edu/publication/c8_date_2017_clock/</link>
      <pubDate>Fri, 24 Mar 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c8_date_2017_clock/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C7] A Low-Power Wireless Image Sensor Node with Noise-Robust Moving Object Detection and a Region-of-Interest Based Rate Controller</title>
      <link>https://iris-lab.skku.edu/publication/c7_gomactech_2017/</link>
      <pubDate>Wed, 22 Mar 2017 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c7_gomactech_2017/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C6] Reconfigurable 96x128 Active Pixel Sensor with 2.1μW/mm2 Power Generation and Regulated Multi- Domain Power Delivery for Self-Powered Imaging</title>
      <link>https://iris-lab.skku.edu/publication/c6_esscirc_2016/</link>
      <pubDate>Fri, 23 Sep 2016 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c6_esscirc_2016/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C5] An Energy-Efficient Wireless Video Sensor Node with a Region-of-Interest Based Multi-Parameter Rate Controller for Moving Object Surveillance</title>
      <link>https://iris-lab.skku.edu/publication/c5_avss_2016/</link>
      <pubDate>Tue, 23 Aug 2016 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c5_avss_2016/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C4] An Energy-Aware Approach to Noise-Tolerant Moving Object Detection for Low-Power Wireless Image Sensor Platforms</title>
      <link>https://iris-lab.skku.edu/publication/c4_islped_2016/</link>
      <pubDate>Mon, 08 Aug 2016 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c4_islped_2016/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C3] A Self-powered Wireless Video Sensor Node for Moving Object Surveillance</title>
      <link>https://iris-lab.skku.edu/publication/c3_gomactech_2016/</link>
      <pubDate>Tue, 22 Mar 2016 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c3_gomactech_2016/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[J1] An Energy-Efficient Wireless Video Sensor Node for Moving Object Surveillance</title>
      <link>https://iris-lab.skku.edu/publication/j1_tmscs_2015/</link>
      <pubDate>Thu, 17 Sep 2015 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/j1_tmscs_2015/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C2] Exploring Power Attack Protection of Resource Constrained Encryption Engines using Integrated Low-Drop-Out Regulators</title>
      <link>https://iris-lab.skku.edu/publication/c2_islped_2015/</link>
      <pubDate>Sat, 08 Aug 2015 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c2_islped_2015/</guid>
      <description></description>
    </item>
    
    <item>
      <title>[C1] Adaptive Wireless Video Sensor Node Using Content-Aware Pre-Processing for Moving Target Identification</title>
      <link>https://iris-lab.skku.edu/publication/c1_gomactech_2015/</link>
      <pubDate>Sun, 22 Mar 2015 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publication/c1_gomactech_2015/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/accomplishments/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/accomplishments/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/contact/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/contact/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/members/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/members/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/news/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/news/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/posts/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/posts/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/professor/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/professor/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/projects/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/projects/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/publications/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/publications/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/recruiting/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/recruiting/</guid>
      <description></description>
    </item>
    
    <item>
      <title></title>
      <link>https://iris-lab.skku.edu/research/</link>
      <pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate>
      <guid>https://iris-lab.skku.edu/research/</guid>
      <description></description>
    </item>
    
  </channel>
</rss>
