I am an associate professor of Artificial Intelligence and Computer Science at POSTECH, and a faculty member of POSTECH Computer Vision Lab. I did a post-doc with Prof. Ivan Laptev and Prof. Jean Ponce in the WILLOW team at the Department of Computer Science of the École Normale Supérieure and Inria Paris. I completed my BS and PhD at POSTECH where I worked with Prof. Joon Hee Han and Prof. Bohyung Han.

My research lies at the intersection of computer vision and machine learning. I am particularly interested in improving robustness and efficiency of physical AI (such as world models and vision-language-action models), integrated and compositional representation of omni-modal data, and state space models for vision.

News

Prospective Students

  • Graduate students: This year, we are planning to recruit graduate students interested in working on world models, vision-language-action models, and multimodal knowledge graphs for agents, exclusively from the Graduate School of AI. To apply for our lab, please submit your information through this Google form. Priority will be given to applicants who have completed an internship within our group or those who intend to pursue a PhD program in our group.
  • Undergrad internship: We are seeking highly motivated undergraduate interns. If you are a POSTECH student, please contact me directly via email. Students from other universities are strongly recommended to apply through the POSTECH SURF or WURF programs.

    • Recent Achievements

  • Compact world models: Check out our latest breakthrough that enables world modeling using only 8 tokens (CompACT). We have drastically increased planning speed by up to 40x while maintaining planning accuracy!
  • Robust vision foundation models: Vision foundation models often fail in adverse weather or low-light conditions. Explore our collaborative research with Google and ETH Zurich aimed at dramatically improving the robustness of these models (GaRA-SAM, RobustPVOS).
  • State space models (SSMs) for vision: Discover our research demonstrating how the compressive properties of SSMs can enhance the neural representation of visual data (Structured State Space Kernel). Stay tuned for more updates on Jinsung’s ongoing work in SSMs for vision!

    • Honors and Awards

  • (2026. 02) Sohyun won Best PhD Dissertation Award from the College of IT, POSTECH.
  • (2026. 02) Dongwon won Best Dissertation Award from the Department of CS, POSTECH.
  • (2026. 02) Jinsung won POSTECHIAN Fellowship.
  • (2026. 02) Our team won Best Paper Award - Gold Prize at IPIU 2026.

    • Selected Publications

    Full publication list is available here.
    CompACT

    Planning in 8 Tokens: A Compact Discrete Tokenizer for Latent World Model

    Dongwon Kim, Gawon Seo, Jinsung Lee, Minsu Cho, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
    Robust Promptable VOS

    Robust Promptable Video Object Segmentation

    Sohyun Lee, Yeho Gwon, Lukas Hoyer, Konrad Schindler, Christos Sakaridis, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
    State-Space Models for Data-specific Neural Representation

    Exploring State-Space Models for Data-Specific Neural Representation

    Jinsung Lee and Suha Kwak
    International Conference on Learning Representations (ICLR), 2026
    GaRA-SAM

    GaRA-SAM: Robustifying Segment Anything Model with Gated-Rank Adaptation

    Sohyun Lee, Yeho Gwon, Lukas Hoyer, and Suha Kwak
    Advances in Neural Information Processing Systems (NeurIPS), 2025
    MemDistill

    MemDistill: Distilling LiDAR Knowledge into Memory for Camera-Only 3D Object Detection

    Donghyeon Kwon, Youngseok Yoon, Hyeongseok Son, and Suha Kwak
    IEEE/CVF International Conference on Computer Vision (ICCV), 2025
    GENIUS

    GENIUS: A Generative Framework for Universal Multimodal Search

    Sungyeon Kim, Xinliang Zhu, Xiaofan Lin, Muhammet Bastan, Douglas Gray, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
    Audio-guided Video Representation

    Learning Audio-guided Video Representation with Gated Attention for Video-Text Retrieval

    Boseung Jeong, Jicheol Park, Sungyeon Kim, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
    Oral Presentation
    Bootstrapping Top-down Information

    Bootstrapping Top-down Information for Self-modulating Slot Attention

    Dongwon Kim, Seoyeon Kim, and Suha Kwak
    Advances in Neural Information Processing Systems (NeurIPS), 2024
    Classification Matters

    Classification Matters: Improving Video Action Detection with Class-Specific Attention

    Jinsung Lee, Taeoh Kim, Inwoong Lee, Minho Shim, Dongyoon Wee, Minsu Cho, and Suha Kwak
    European Conference on Computer Vision (ECCV), 2024
    Oral Presentation
    Efficient and Versatile Robust Fine-Tuning

    Efficient and Versatile Robust Fine-Tuning of Zero-shot Models

    Sungyeon Kim, Boseung Jeong, Donghyun Kim, and Suha Kwak
    European Conference on Computer Vision (ECCV), 2024
    Qualcomm Innovation Fellowship Winner
    Online Temporal Action Localization

    Online Temporal Action Localization with Memory-augmented Transformer

    Youngkil Song*, Dongkeun Kim*, Minsu Cho, and Suha Kwak (*equal contribution)
    European Conference on Computer Vision (ECCV), 2024
    FREST

    FREST: Feature RESToration for Semantic Segmentation under Multiple Adverse Conditions

    Sohyun Lee, Namyup Kim, Sungyeon Kim, and Suha Kwak
    European Conference on Computer Vision (ECCV), 2024
    Group Activity Detection

    Towards More Practical Group Activity Detection: A New Benchmark and Model

    Dongkeun Kim, Youngkil Song, Minsu Cho, and Suha Kwak
    European Conference on Computer Vision (ECCV), 2024
    Extreme Point Instance Segmentation

    Extreme Point Supervised Instance Segmentation

    Hyeonjun Lee, Sehyun Hwang, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024
    Active Learning for Semantic Segmentation

    Active Learning for Semantic Segmentation with Multi-class Label Query

    Sehyun Hwang, Sohyun Lee, Hoyoung Kim, Minhyeon Oh, Jungseul Ok, and Suha Kwak
    Advances in Neural Information Processing Systems (NeurIPS), 2023
    Shatter and Gather

    Shatter and Gather: Learning Referring Image Segmentation with Text Supervision

    Dongwon Kim*, Namyup Kim*, Cuiling Lan, and Suha Kwak (*equal contribution)
    IEEE/CVF International Conference on Computer Vision (ICCV), 2023
    PromptStyler

    PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization

    Junhyeong Cho, Gilhyun Nam, Sungyeon Kim, Hunmin Yang, and Suha Kwak
    IEEE/CVF International Conference on Computer Vision (ICCV), 2023
    Proxy TTA

    Leveraging Proxy of Training Data for Test-Time Adaptation

    Juwon Kang, Nayeong Kim, Donghyeon Kwon, Jungseul Ok, and Suha Kwak
    International Conference on Machine Learning (ICML), 2023
    Qualcomm Innovation Fellowship Winner
    Set Embedding

    Improving Cross-Modal Retrieval with Set of Diverse Embeddings

    Dongwon Kim, Namyup Kim, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023
    Highlight
    HIER

    HIER: Metric Learning Beyond Class Labels via Hierarchical Regularization

    Sungyeon Kim, Boseung Jeong, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023
    ExLPose

    Human Pose Estimation in Extremely Low-Light Conditions

    Sohyun Lee*, Jaesung Rim*, Boseung Jeong, Geonu Kim, Byungju Woo, Haechan Lee, Sunghyun Cho, and Suha Kwak (*equal contribution)
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023
    WEDGE

    WEDGE: Web-Image Assisted Domain Generalization for Semantic Segmentation

    Namyup Kim, Taeyoung Son, Jaehyun Pahk, Cuiling Lan, Wenjun Zeng, and Suha Kwak
    IEEE International Conference on Robotics and Automation (ICRA), 2023
    LAMDA

    Combating Label Distribution Shift for Active Domain Adaptation

    Sehyun Hwang, Sohyun Lee, Sungyeon Kim, Jungseul Ok, and Suha Kwak
    European Conference on Computer Vision (ECCV), 2022
    Qualcomm Innovation Fellowship Winner
    FIFO

    FIFO: Learning Fog-invariant Features for Foggy Scene Segmentation

    Sohyun Lee, Taeyoung Son, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022
    Best Paper Finalist, Oral Presentation, Qualcomm Innovation Fellowship Winner
    ReSTR

    ReSTR: Convolution-free Referring Image Segmentation Using Transformers

    Namyup Kim, Dongwon Kim, Cuiling Lan, Wenjun Zeng, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022
    Qualcomm Innovation Fellowship Winner
    STML

    Self-Taught Metric Learning without Labels

    Sungyeon Kim, Dongwon Kim, Minsu Cho, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022
    Qualcomm Innovation Fellowship Winner
    ELN

    Semi-supervised Semantic Segmentation with Error Localization Network

    Donghyeon Kwon and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022
    Style Neophile

    Style Neophile: Constantly Seeking Novel Styles for Domain Generalization

    Juwon Kang, Sohyun Lee, Namyup Kim, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022
    Qualcomm Innovation Fellowship Winner
    URIE

    URIE: Universal Image Enhancement for Visual Recognition in the Wild

    Taeyoung Son, Juwon Kang, Namyup Kim, Sunghyun Cho, and Suha Kwak
    European Conference on Computer Vision (ECCV), 2020
    Proxy Anchor Loss

    Proxy Anchor Loss for Deep Metric Learning

    Sungyeon Kim, Dongwon Kim, Minsu Cho, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020
    Continuous Metric Learning

    Deep Metric Learning Beyond Binary Supervision

    Sungyeon Kim, Minkyo Seo, Ivan Laptev, Minsu Cho, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
    Oral Presentation, Qualcomm Innovation Fellowship Winner
    Weakly Supervised Instance Segmentation

    Weakly Supervised Learning of Instance Segmentation with Inter-pixel Relations

    Jiwoon Ahn, Sunghyun Cho, and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
    Oral Presentation
    Weakly Supervised Semantic Segmentation

    Learning Pixel-level Semantic Affinity with Image-level Supervision for Weakly Supervised Semantic Segmentation

    Jiwoon Ahn and Suha Kwak
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2018
    Winner of the KCCV Sang-Uk Lee Prize

    Professional Service

     Editorial Board Member
    • International Journal of Computer Vision (IJCV), 2021 ~ 2024
     Area Chair or Equivalent
    • IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022 ~ 2026
    • IEEE/CVF International Conference on Computer Vision (ICCV), 2019 ~ 2025
    • European Conference on Computer Vision (ECCV), 2024
    • Advances in Neural Information Processing Systems (NeurIPS), 2023 ~ 2026
    • International Conference on Learning Representations (ICLR), 2025, 2026
    • International Conference on Machine Learning (ICML), 2025, 2026
    • AAAI Conference on Artificial Intelligence (AAAI), 2024, 2026
    • IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023 ~ 2026
    • Asian Conference on Computer Vision (ACCV), 2022 ~ 2024

    Contact

    +82-54-279-2390
    315 B2, POSTECH

    77 Cheongam-ro, Nam-gu
    Pohang, Gyungbuk
    37673 Korea

    suha.kwak (a) postech.ac.kr