Junwon Seo

I am a PhD student in Robotics at Carnegie Mellon University (CMU). My research focuses on enabling robots to operate safely and reliably under uncertainty. Before joining CMU, I was a research officer at the Defense AI Center of the Agency for Defense Development (ADD), where I researched uncertainty-aware navigation in complex and unknown environments. I completed my bachelor's degree (Summa Cum Laude) in Computer Science and Engineering at the Seoul National University (SNU).

Email  |  CV  |  Google Scholar  |  YouTube  |  LinkedIn

headshot
Research Interests

My research goal is to enable robots to navigate safely and effectively in uncertain environments. I am interested in capturing the uncertainty in learning-based models and adapting these models to enhance safety and robustness. To this end, I am currently focusing on the following topics, but not limited to:

  • Uncertainty-aware Perception
  • Uncertainty Quantification
  • Robotic Mapping
  • Meta Learning

    		•
    Publications
    (Equal contributions are denoted by *)
    Uncertainty-aware Semantic Mapping in Off-road Environments with Dempster-Shafer Theory of Evidence
    Junyoung Kim, Junwon Seo
    ICRA 2024 Workshop on Resilient Off-road Autonomy, 2024.
    project page | paper

    Robust semantic mapping with Evidential Deep Learning and Dempster-Shafer Theory of Evidence.

    Evidential Semantic Mapping in Off-road Environments with Uncertainty-aware Bayesian Kernel Inference
    Junyoung Kim*, Junwon Seo*, Jihong Min
    IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024.
    project page | paper | video

    Uncertainty-aware semantic BKI mapping framework for robust deployments in off-road environments using Evidential Deep Learning.

    METAVerse: Meta-Learning Traversability Cost Map for Off-Road Navigation
    Junwon Seo, Taekyung Kim, Seongyong Ahn, Kiho Kwak
    IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024.
    paper | video

    Meta-learning framework for learning a global model that accurately and reliably predicts off-road terrain traversability in various environments.

    Scalable Off-road Semantic Terrain Map Estimation using Uncertainty-aware LiDAR-Camera Fusion
    Ohn Kim*, Junwon Seo*, Seongyong Ahn
    IEEE Intelligent Vehicles Symposium (IV) Workshop on Off-road Autonomy, 2024.
    paper

    UFO: Uncertainty-aware LiDAR-image Fusion for Off-road Semantic Terrain Map Estimation

    DA-RAW: Domain Adaptive Object Detection for Real-World Adverse Weather Conditions
    Minsik Jeon*, Junwon Seo*, Jihong Min
    IEEE International Conference on Robotics and Automation (ICRA), 2024.
    paper | video

    Unsupervised domain adaptation for robust object detection in real-world adverse weather conditions.

    Bridging Active Exploration and Uncertainty-Aware Deployment Using Probabilistic Ensemble Neural Network Dynamics
    Taekyung Kim*, Jungwi Mun*, Junwon Seo, Beomsu Kim, Seongil Hong
    Robotics: Science and Systems (RSS), 2023.
    paper | project page | video

    Active exploration for learning vehicle dynamics model and uncertainty-aware deployment.

    Safe Navigation in Unstructured Environments by Minimizing Uncertainty in Control and Perception
    Junwon Seo, Jungwi Mun, Taekyung Kim
    RSS Workshop on Inference and Decision Making for Autonomous Vehicles (IDMAV), 2023.
    paper

    Closed-loop integration of perception and control methods, which minimize uncertainty for safe and efficient off-road navigation.

    Learning Off-Road Terrain Traversability with Self-Supervisions Only
    Junwon Seo, Sungdae Sim, Inwook Shim
    IEEE Robotics and Automation Letters (RA-L), 2023.
    paper | video

    Visual terrain traversability learning in off-road that utilizes only self-supervision without human annotations.

    Scalable Framework for Self-Supervised Traversability Estimation in Unstructured Environments
    Junwon Seo*, Taekyung Kim*, Kiho Kwak, Jihong Min, Inwook Shim
    IEEE Robotics and Automation Letters (RA-L), 2023.
    paper | video | presentation

    LiDAR-based self-supervised traversability estimation framework that can be easily scalable to various vehicles and sensor types.

    Self-Supervised 3D Traversability Estimation with Proxy Bank Guidance
    Jihwan Bae*, Junwon Seo*, Taekyung Kim, Hae-gon Jeon, Kiho Kwak, Inwook Shim
    IEEE Access, 2023.
    paper

    Handling uncertainty problems in self-supervised traversability estimation with metric learning.

    Projects


    Adaptive Path Planning Based on Situational Awareness and Dynamic Model Learning
    Agency for Defense Development, 2021 - Present

    Perception and control algorithms for off-road navigation: learning-based traversability cost map, semantic terrain classification map, moving object detection and tracking, and vehicle dynamics learning.

    Traversable Area and Object Detection on Adverse Environmental Condition
    Agency for Defense Development, 2021 - 2022

    Built an autonomous driving dataset with multi-spectral images and LiDAR in adverse weather conditions and developed a robust traversable area detection and object detection algorithm.

    Autonomous Tunnel Exploitation
    Agency for Defense Development, 2021 - 2022

    Developed a learning-based traversability map for autonomous robotic exploration under subterranean environments.

    3D Hair Styling Simulation
    Software Maestro, 2020

    A deep-learning-based algorithm for face and hair reconstruction from a single image and a virtual 3D hairstyle simulation service.

    		Light Stage
    Morpehus 3D, 2020

    The light stage system and face normal map reconstruction algorithm for 3D face reconstruction with high fidelity.

    Structured-Light 3D Hair and Face Scanner
    Morpheus 3D, 2019 - 2020

    A Structured-light 3D scanner for high-quality hair and face geometry reconstruction.


    template adapted from this awesome website