Junghyup Lee

Joint MS/PhD student in Computer Vision Lab @ Yonsei University.

About

I am a Ph.D. candidate in School of Electrical and Electronic Engineering at Yonsei University.

My research interests span a wide spectrum of machine learning and computer vision, mainly focusing on efficient machine learning (network quantization, network architecture search, knowledge distillation), image matching (semantic correspondence), and image restoration (super-resolution).

Education

  • Joint course of M.S./Ph.D degrees in School of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea (Mar. 2018 - Aug. 2024 (expected)).
  • B.S. in School of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea (Mar. 2012 - Feb. 2018).

Publications

First author

  • AZ-NAS: Assembling Zero-Cost Proxies for Network Architecture Search
    Junghyup Lee and Bumsub Ham.
    in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun. 2024.
    [Paper] [Project Page]
  • Learning Semantic Correspondence Exploiting an Object-level Prior
    Junghyup Lee*, Dohyung Kim*, Wonkyung Lee, and Bumsub Ham (*equal contribution).
    IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), vol. 44, no. 3, pp. 1399-1414, Mar. 2022.
    [Paper] [Project Page]
  • Network Quantization with Element-wise Gradient Scaling
    Junghyup Lee, Dohyung Kim, and Bumsub Ham.
    in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun. 2021.
    [Paper] [Project Page]
  • Learning with Privileged Information for Efficient Image Super-Resolution
    Wonkyung Lee*, Junghyup Lee*, Dohyung Kim*, and Bumsub Ham (*equal contribution).
    in European Conference on Computer Vision (ECCV), Aug. 2020.
    [Paper] [Project Page]
  • SFNet: Learning Object-aware Semantic Correspondence
    Junghyup Lee*, Dohyung Kim*, and Bumsub Ham (*equal contribution).
    in IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (Oral Presentation), Jun. 2019.
    [Paper] [Project Page]

Co-author

  • PLoPS: Localization-Aware Person Search with Prototypical Normalization
    Sanghoon Lee, Youngmin Oh, Donghyeon Baek, Junghyup Lee, and Bumsub Ham.
    Pattern Recognition (PR), Apr. 2024. (Accepted)
    [Paper]
  • RankMixup: Ranking-Based Mixup Training for Network Calibration
    Jongyoun Noh, Hyekang Park, Junghyup Lee, and Bumsub Ham.
    in International Conference on Computer Vision (ICCV), Oct. 2023.
    [Paper] [Project Page]
  • Decomposed Knowledge Distillation for Class-Incremental Semantic Segmentation
    Donghyeon Baek, Youngmin Oh, Sanghoon Lee, Junghyup Lee, and Bumsub Ham.
    in Neural Information Processing Systems (NeurIPS), Nov. 2022.
    [Paper] [Project Page]
  • SIF-NPU: A 28nm 3.48 TOPS/W 0.25 TOPS/mm2 CNN Accelerator with Spatially Independent Fusion for Real-Time UHD Super-Resolution
    Sumin Lee, Ki-Beom Lee, Sunghwan Joo, Hong Keun Ahn, Junghyup Lee, Dohyung Kim, Bumsub Ham, and Seong-Ook Jung.
    in IEEE European Solid State Circuits Conference (ESSCIRC), Sep. 2022.
    [Paper]
  • OIMNet++: Prototypical Normalization and Localization-aware Learning for Person Search
    Sanghoon Lee, Youngmin Oh, Donghyeon Baek, Junghyup Lee, and Bumsub Ham.
    in European Conference on Computer Vision (ECCV), Oct. 2022.
    [Paper] [Project Page]
  • Learning by Aligning: Visible-Infrared Person Re-identification using Cross-Modal Correspondences
    Hyunjong Park*, Sanghoon Lee*, Junghyup Lee, and Bumsub Ham (*equal contribution).
    in International Conference on Computer Vision (ICCV), Oct. 2021.
    [Paper] [Project Page]
  • Video-based Person Re-identification with Spatial and Temporal Memory Networks
    Chanho Eom, Geon Lee, Junghyup Lee, and Bumsub Ham.
    in International Conference on Computer Vision (ICCV), Oct. 2021.
    [Paper] [Project Page]
  • Distance-aware Quantization
    Dohyung Kim, Junghyup Lee, and Bumsub Ham.
    in International Conference on Computer Vision (ICCV), Oct. 2021.
    [Paper] [Project Page]

Preprint

  • Transition Rate Scheduling for Quantization-Aware Training
    Junghyup Lee, Dohyung Kim, Jeimin Jeon, and Bumsub Ham.
    ArXiv preprint.
    [Paper]

Awards

Scholarship & Funding

  • Global Ph.D Fellowship from National Research Foundation of Korea (NRF), 2019-2023.
  • Yangyoung Scholarship provided by Yangyoung Foundataion, 2016-2017.

Domestic Conferences & Invited Talks

  • Invited Talk (Network Quantization)
    Naver Labs Seminar, Jun. 2022.
  • Poster Presentation (Network Quantization with Element-wise Gradient Scaling)
    Naver AI Author Meetup - Computer Vision, Sep. 2021.
    Korean Conference on Computer Vision (KCCV), Aug. 2021.
  • Poster Presentation (시멘틱 정합을 이용한 비디오 포즈 전파)
    32nd Workshop on Image Processing and Image Understanding (IPIU), Jan. 2020.
  • Poster Presentation (SFNet: Learning Object-aware Semantic Correspondence)
    Samsung AI Forum (SAIF), Nov. 2019.
    Workshop on Frontiers of Electrical Engineering (FREE) in Yonsei University, Oct. 2019.
    Korean Conference on Computer Vision (KCCV), Jul. 2019.

Patents

Registration

  • 시공간 메모리 네트워크를 활용한 동영상 기반 사람 재식별 장치 및 방법
    Apparatus and Method for Person Re-Identification based on Video with Spatial and Temporal Memory Networks
    10-2678912, KR, Jun. 2024.
  • 사전 정보 학습 기반 영상 업스케일링 장치 및 방법
    Image Upscaling Apparatus and Method Based on Learning with Privileged Information
    10-2543690, KR, Jun. 2023.
  • 인공 신경망을 위한 양자화기 및 이의 손실 역전파 방법
    Quantizer for Artificial Neural Networks and Loss Backpropagation Method Thereof
    10-2409476, KR, Jun. 2022.
  • 시멘틱 매칭 장치 및 방법
    Semantic Matching Apparatus and Method
    10-2166117, KR, Oct. 2020.

Application

  • 네트워크 구조 탐색을 위한 비학습 인공지능 신경망 평가 장치 및 방법
    Training-Free Evaluation Apparatus and Method of Artificial Neural Network for Network Architecture Search
    10-2024-0078638, KR, Jun. 2024.
  • 다중 믹스업 기반 딥러닝 네트워크의 신뢰도 교정 장치 및 방법
    Apparatus and Method for Calibrating Confidence of Deep Learning Network Based on Multiple Mixup
    10-2024-0034262, KR, Mar. 2024.
  • 인공 신경망을 위한 양자화 장치 및 방법
    Quantization Apparatus and Method for Artificial Neural Network
    10-2023-0116857, KR, Sep. 2023.
  • 양자화 학습 장치 및 방법
    Quantization-Aware Training Apparatus and Method
    10-2023-0049837, KR, Apr. 2023.
  • 분해 지식 증류 기반 클래스 증분 시멘틱 분할 학습 장치 및 방법
    Apparatus and Method for Class Incremental Semantic Segmentation Learning based on Decomposed Knowledge Distillation
    10-2022-0185609, KR, Dec. 2022.
  • 인공 신경망을 위한 양자화기 및 양자화 방법
    Quantizer and Quantization Method for Artificial Neural Network
    10-2020-0135673, KR, Oct. 2020.

Experiences

  • Reviewer
    IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020, 2021, 2022, 2023, 2024.
    Neural Information Processing Systems (NeurIPS), 2020, 2021, 2022.
    International Conference on Computer Vision (ICCV), 2019, 2021.
    European Conference on Computer Vision (ECCV), 2020.
    AAAI Conference on Artificial Intelligence (AAAI), 2023.
    IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI).

  • Teaching Assistant in School of Electrical and Electronic Engineering, Yonsei University
    Digital Image Processing (EEE5320), 2021-01.
    Random Process (EEE5110), 2020-01.
    Signal Processing Lab. (EEE4423), 2020-01.
    Smart Technology AIR (ENG2116), 2019-01.
    Signal Processing Lab. (EEE4423), 2019-01.
    Electrical and Electronic Engineering Experiments: Applications (EEE3313), 2018-02.
    Data Structure (EEE2020), 2018-01.