Education

• Ph.D., School of Electrical Engineering, KAIST                                                                                  2020.03 - 2024.08 (expected)

        Advisor : Prof. Changick Kim

• M.S., School of Electrical Engineering, KAIST                                                                                    2018.03 - 2020.02

        Advisor : Prof. Changick Kim

Thesis: A Two-Branch Feature Selective Network Exploiting Class-Activated Regions for Clothes Recognition

• B.S., School of Electronics Engineering, Kyungpook National University                                    2013.03 - 2018.02

        GAP : 4.14 / 4.3

Research Interest

• Deep Learning, Computer Vision

• Video understanding : Action recognition, Group Activity Recognition, Action localization, Action Anticipation, etc.

• Multi-modal learning 

Journal

• • Sangmin Woo, Jinyoung Park, Inyong Koo, Sumin Lee, Minki Jeong, and Changick Kim, "Explore-And-Match: Bridging Proposal-Based and Proposal-Free with Transformer for Sentence Grounding in Videos," Submitted to Multimedia Tools and Applications, Nov. 2023. [paper]

  • Sumin Lee, Sangmin Woo, Muhammad Adi Nugroho, and Changick Kim, "Modality Mixer Exploiting Complementary Information for Multi-modal Action Recognition," In Major Revision, IEEE Transactions on Image Processing (TIP), Jul. 2023. [paper]

  • Yeonju Park, Sangmin Woo, Sumin Lee, Muhammad Adi Nugroho, and Changick Kim, "Cross-Modal Alignment and Translation for Missing Modality Action Recognition," Computer Vision and Image Understanding (CVIU), vol. 236, pages 103805-130813, Aug. 2023. [SCI, IF:4.5] [paper]

  •  Sumin Lee, Hyunjun Eun, Jinyoung Moon, Seokeon Choi, Yoonhyung Kim, Chanho Jung, and Changick Kim, "Learning to Discriminate Information for Online Action Detection: Analysis and Application", IEEE Transaction on Pattern Analysis and Machine Intelligence (TPAMI), vol. 45, no. 5, pp. 5918-5934, 1 May 2023. [SCI, IF:23.6] [paper]

  • Hyunjun Eun, Sumin Lee, J. Moon, J. Park, Chanho Jung, and Changick Kim, "SRG: Snippet Relatedness-based Temporal Action Proposal Generator," IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), vol. 30, no. 11, pp.  4232-4244, Nov. 2020. [SCI, IF : 8.4]  [paper]

  • Sumin Lee, Hyunjun Eun, Sungchan Oh, Wonjun Kim, Chanho Jung, and Changick Kim, "Landmark-free Clothes Recognition with a Two-Branch Feature Selective Network,"  Electronics Letters, vol. 55, no. 13, pp. 745-747,  June 2019.  [paper]

Conference

• • Sumin Lee, Yooseung, Sangmin Woo, and Changick Kim, "Spatio-Temporal Proximity-Aware Dual-Path Model for Panoramic Activity Recognition", Arxiv 2024. [paper]

  • Sangmin Woo, So-Yeong Jeon, Jinyoung Park, Minji Son, Sumin Lee, and Changick Kim, "Sketch-based Video Object Localization," in Proc. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV),  Hawaii, USA, Jan. 4-8, 2024. [paper]

  • Donguk Kim, Sumin Lee, Sangmin Woo, Jinyoung Park, Muhammad Adi Nugroho,  and Changick Kim, "Multi-modal Social Group Activity Recognition in Panoramic Scene," in Proc. IEEE International Conference on Visual Communications and Image Processing (VCIP), Jeju, Korea, Dec. 4-6, 2023. [paper]

  • Muhammad Adi Nugroho, Sangmin Woo, Sumin Lee, and Changick Kim,"AHFu-Net: Align, Hallucinate, and Fuse Network for Missing Multimodal Action Recognition," in Proc. IEEE International Conference on Visual Communications and Image Processing (VCIP), Jeju, Korea, Dec. 4-6, 2023. (Oral) [paper]

  • Muhammad Adi Nugroho, Sangmin Woo, Sumin Lee, and Changick Kim, "Audio-Visual Glance Network for Efficient Video Recognition," in Proc. IEEE International Conference on Computer Vision (ICCV), Paris, France, Oct. 2-6, 2023. [paper]

  • Sangmin Woo, Sumin Lee, Yeonju Park, Muhammad Adi Nugroho, and Changick Kim, "Towards Good Practices for Missing Modality Robust Action Recognition," in Proc. the Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI), Washington DC, USA, Feb. 7-14, 2023. (Oral) [paper]

  • Sumin Lee, Sangmin Woo, Yeonju Park, Muhammad Adi Nugroho, and Changick Kim, "Modality Mixer for Multi-modal Action Recognition," in Proc. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Hawaii, USA, Jan. 3-7, 2023. [paper]

  • Jinyoung Park, Minseok Son, Sumin Lee, and Changick Kim, "DAT: Domain Adaptive Transformer for Domain Adaptive Semantic Segmentation ," in Proc. the 29th IEEE International Conference on Image Processing (ICIP), Bordeaux, France, Oct. 16-19, 2022. [paper]

  • Seokeon Choi, Sumin Lee, Youngeun Kim, Taekyung Kim, Changick Kim, "Hi-CMD: Hierarchical Cross-Modality Disentanglement for Visible-Infrared Person Re-Identification ," in Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, Washington, USA, Jun. 14-19, 2020. [paper][github]

  • Sumin Lee, Sungchan Oh, Chanho Jung, and Changick Kim, "A Global-local Embedding Module for Fashion Landmark Detection," in Proc. ICCV 2019 Workshop, Seoul, Korea, Oct. 27 - Nov. 2, 2019.  [paper][github]

Domestic Papers

• • 이수민, 김창익, "멀티모달 행동 인식을 위한 적대적 보강 학습 방법에 대한 연구," 한국군사과학기술학회 종합학술대회, pp. 274-275, 제주국제컨벤션센터(ICC) 제주, Jun. 15-16, 2023.

  • 이수민, 오성찬, 정찬호, 김창익, "딥러닝을 이용한 시각적 의류 분석 기술: 서베이," 한국통신학회논문지, vol.45, no. 7, pp. 1174-1182, 2020.

  • 이수민, 오성찬, 정찬호, 김창익, "의류 인식을 위한 의류 랜드마크 특징 탐색 네트워크," 한국통신학회논문지, vol. 45, no. 5, pp. 801-804, 2020.

  • 이수민, 오성찬, 정찬호, 김창익, "의류 인식을 위한 다중 도메인 학습 기법," 제29회 신호처리합동학술대회, pp. 196-198, KAIST 대전, Sep. 26-27, 2019.

  • 이수민, 은현준, 오성찬, 정찬호, 김창익, " 두 개 분기 네트워크를 이용한 약지도 의류 인식," 대한전자공학회 추계학술대회, 송도컨벤시아, Nov.23, 2018

• 2022.01 ~ 2024.06
  Agency for Defense Development (ADD)
  Multi-modal human activity recognition
          National Defense AI Specialized Research Center 

• 2021.03 ~ 2021.11
          Electronics and Telecommunications Research Institute (ETRI)
          Title : Spatio-Temporal Action Localization in Video (비디오 구간 예측을 위한 행동 모델링 고도화 연구)

• 2020.03 ~ 2020.11
  Electronics and Telecommunications Research Institute (ETRI)
  Title : Action Prediction (시각 기억 기반의 행동 예측 기술 연구)

• 2019.05 ~ 2019.11
  Electronics and Telecommunications Research Institute (ETRI)
  Title : Visual object detection and attribute recognition (시각 특성 영역 검출 및 인식 기술 개발)

• 2018.05 ~ 2018.11
  Electronics and Telecommunications Research Institute (ETRI)
  Title : Fashion image based attribute extraction (영상기반 특성정보 추출) 

Conference Reviewer

• IEEE International Conference on Image Processing (ICIP)

• IEEE International Conference on Visual Communications and Image Processing (VCIP)

• IEEE Winter Conference on Applications of Computer Vision (WACV)

Journal Reviewer

•  IEEE Transactions on Circuits and Systems for Video Technology (TCSVT)

• Changick Kim, Sumin Lee, Jinyoung Park, Donguk Kim, Muhammad Adi Nugroho, Sangmin Woo, Yooseung Wang,  "Method for group activity recognition using RGB videos and LiDAR data". Application number: 10-2024-0034087, Application date: 2024-03-11.

• Changick Kim, Inyong Koo, Yeonju Park, Minki Jeong, Sumin Lee, Sangmin Woo, "Method and appratus for human activity recognition using accelerometer and gyroscope sensors". Registration number: 10-2515065.  Registration date: 2023.03.23.

• Representative of Computational Intelligence Lab.
  Sep. 2022 - Oct. 2023

• Presidential Commendation Award of KAIST (Jan. 2023)
  KAIST Center of Applied Research in Artificial Intelligent

• 2020 AI City Challenge Track 3 "City-Scale Multi-Camera Vehicle Tracking" : 3rd Prize (Apr. 2020)

• Paper Competition Awards, Qualcomm-KAIST Innovation Awards 2019
  [Paper] Hyunjun Eun, Sumin Lee, J. Moon, J. Park, Chanho Jung, Changick Kim, "SRG: Snippet Relatedness-based Temporal Action Proposal Generator"

• LG Electronics Paper Awards, Nov. 23, 2018, Autumn Annual Conference of IEIE
  [Paper]  Sumin Lee, Hyunjun Eun, Sungchang Oh, Chanho Jung, and Changick Kim, "Weakly Supervised Clothes Recognition with a Two-Branch Network"