Seokha Moon

shmoon96[at]korea[dot]ac[dot]kr

prof_pic.jpg

Hi!😀 I’m a Ph.D. Student at Korea University, Vision & AI lab (Advisor: Prof. Jinkyu Kim). I earned a Bachelor’s degree in Computer Science from Yonsei University.

My research interests lie in the field of Autonomous Robots 🤖 and Autonomous Driving 🚗. Recently, I have been focusing on camera-based perception tasks in autonomous driving, including 3D detection and occupancy prediction. I also explored trajectory prediction with an emphasis on modeling interactions between agents, using vision-driven text guidance as supervision to inform the model of relevant contextual cues needed to understand each agent’s situation.

Currently, I am particularly interested in Vision-Language Navigation for autonomous robots and End-to-End frameworks for autonomous driving.

news

Mar 28, 2025 💼📄 Our paper “Mitigating Trade-off: Stream and Query-guided Aggregation for Efficient and Effective 3D Occupancy Prediction” has been released on arXiv! The work was done during my internship at NAVER LABS.
Sep 02, 2024 💼 Started a research internship at NAVER LABS, focusing on real-time 3D occupancy prediction for autonomous driving.
Aug 06, 2024 📄✨ Our paper “Who Should Have Been Focused: Transferring Attention-based Knowledge from Future Observations for Trajectory Prediction” has been accepted to ICPR 2024!
Jul 04, 2024 📄✨ Our paper “VisionTrap: Vision-Augmented Trajectory Prediction Guided by Textual Descriptions” has been accepted to ECCV 2024!
Jan 29, 2024 📄✨ Our paper “Learning Temporal Cues by Predicting Objects Move for Multi-camera 3D Object Detection” has been accepted to ICRA 2024!

selected publications

  1. Preprint
    STREAMOCC.png
    Mitigating Trade-off: Stream and Query-guided Aggregation for Efficient and Effective 3D Occupancy Prediction
    Seokha Moon, Janghyun Baek, Giseop Kim, Jinkyu Kim, and Sunwook Choi
    Arxiv Preprint, 2025
  2. ECCV
    VISIONTRAP.png
    VisionTrap: Vision-Augmented Trajectory Prediction Guided by Textual Descriptions
    Seokha Moon, Hyun Woo, Hongbeen Park, Haeji Jung, Reza Mahjourian, Hyung-gun Chi, Hyerin Lim, Sangpil Kim, and Jinkyu Kim
    European Conference on Computer Vision (ECCV), 2024
  3. ICPR
    WHO.png
    Who Should Have Been Focused: Transferring Attention-based Knowledge from Future Observations for Trajectory Prediction
    Seokha Moon, Kyuhwan Yeon, Hayoung Kim, Seong-Gyun Jeong, and Jinkyu Kim
    International Conference on Pattern Recognition (ICPR), 2024
  4. ICRA
    DAP.png
    Learning Temporal Cues by Predicting Objects Move for Multi-camera 3D Object Detection
    Seokha Moon, Hongbeen Park, Jungphil Kwon, Jaekoo Lee, and Jinkyu Kim
    International Conference on Robotics and Automation (ICRA), 2024
  5. WACV
    BEVMAP.png
    BEVMap: Map-Aware BEV Modeling for 3D Perception
    Mincheol Chang, Seokha Moon, Reza Mahjourian, and Jinkyu Kim
    Winter Conference on Applications of Computer Vision (WACV), 2024
  6. BMVC
    ORA3D.png
    ORA3D: Overlap Region Aware Multi-view 3D Object Detection
    Wonseok Roh, Gyusam Chang, Seokha Moon, Giljoo Nam, Chanyoung Kim, Younghyun Kim, Jinkyu Kim, and Sangpil Kim
    British Machine Vision Conference (BMVC), 2022