Embedded Files
Research Topics
Multimodal Learning
Learns unified representations from vision, language, audio, and tactile signals for robust perception and reasoning.
Domain Adaptation
Adapts models to distribution shifts and new environments while preserving reliable performance.
Efficient Machine Learning
Improves learning and inference efficiency while reducing data, compute, and memory requirements.
Video Understanding
Understands objects, actions, events, and temporal relationships in complex videos.
Robot Learning
Builds embodied intelligence through visual, tactile, and multimodal perception for physical interaction.
9
Ongoing Projects
16
Research Members
33
Journal Publications
45
Conference Publications
Latest News
[Project] 2026.09 홍성은 교수님이 교육연구단장으로 참여한 한국연구재단(NRF) 4단계 BK21 사업 「AI+X 융합형 교육연구단 지원 시범사업」이 선정되었습니다.
[Project] 2026.09 김홍엽 학생의 「정보이론 기반 장기 옴니모달 비디오 시공간 토큰 절감 프레임워크」가 한국연구재단 박사과정생연구장려금지원사업에 선정되었습니다.
[Member] 2026.08 김대용 학생과 김승규 학생의 석박사통합과정 합류를 축하합니다.
[Paper] 2026.08 연구실 논문 2편이 EMNLP 2026 Findings에 채택되었습니다: “Block Recursive Transformers for Structured Parameter Sharing”와 “SCENE: A Scenario Completion Benchmark for Measuring Social Bias in Large Language Models”.
[Paper] 2026.08 김재익, 백윤재, 정애천, 김홍엽, 최소윤, 김영은, 도재영 학생이 참여한 “Dual Path Semantic Segmentation with Locality Guidance Tokens” 논문이 Neurocomputing에 게재 승인되었습니다.
[Event] 2026.08 홍성은 교수님이 오거나이저로 참여한 “RespMultimodal 2026: Responsible Multimodal Foundation Models for Knowledge Discovery” 워크숍이 KDD 2026에서 개최되었습니다.
AI & Media Lab (AIM Lab)
Office : 90313, International Hall, Sungkyunkwan University,
25-2 Sungkyunkwan-ro, Jongno-gu, Seoul 03063, Korea
Email : csehong@skku.edu Web : aim.skku.edu Tel : 02-740-1809
Page updated
Google Sites
Report abuse