Xiongkun Linghu

personal_img.jpg

I am now a research engineer at Beijing Institute for General Artificial Intelligence(BIGAI). I obtained my M.S. from Tsinghua University in July 2023. Previously, I received my B.S. from Beijing Institute of Technology in July 2020.

I am generally interested in multimodal foundation models and embodied AI. My long-term goal is building powerful, reliable and safe embodied agents in the digital and physical world.

Publications

  1. scenecot.png
    SceneCOT: Eliciting Grounded Chain-of-Thought Reasoning in 3D Scenes
    Xiongkun Linghu, Jiangyong Huang, Ziyu Zhu, Baoxiong Jia, and Siyuan Huang
    ICLR 2026
    arXiv HTML Code
  2. beacon3d.png
    Unveiling the Mist over 3D Vision-Language Understanding: Object-centric Evaluation with Chain-of-Analysis
    Jiangyong Huang* ,Baoxiong Jia*, Yan Wang, Ziyu Zhu, Xiongkun Linghu, Qing Li, Songchun Zhu, and Siyuan Huang
    CVPR 2025
    arXiv HTML Code
  3. msr3d.png
    Multi-modal Situated Reasoning in 3D Scenes
    Xiongkun Linghu*, Jiangyong Huang* , Xuesong Niu*, Xiaojian Ma, Baoxiong Jia, and Siyuan Huang
    NeurIPS, Datasets and Benchmarks Track, 2024
    arXiv HTML Code
  4. LEO.png
    An Embodied Generalist Agent in 3D World
    Jiangyong Huang*, Silong Yong*, Xiaojian Ma*, Xiongkun Linghu*, Puhao Li, Yan Wang, Qing Li, Songchun Zhu, Baoxiong Jia, and Siyuan Huang
    ICML, 2024
    arXiv HTML Code Demo
  5. SFSC.png
    Switchable representation learning framework with self-compatibility
    Shengsen Wu, Yan Bai, Yihang Lou, Xiongkun Linghu, Jianzhong He, and Ling-Yu Duan
    CVPR, 2023
    arXiv

Preprint

  1. bel.jpg
    Bayesian Evidential Learning for Few-Shot Classification
    Xiongkun Linghu, Yan Bai, Yihang Lou, Shengsen Wu,Jinze Li, Jianzhong He, and Tao Bai
    arxiv, 2022
    arXiv

Service

    • Reviewers: I serve as the reviewer for NeurIPS, ICML, ICLR, CVPR, and ECCV

Experience

    • 2021.12 - 2022.7, Intern, Huawei, Few-shot Learning and uncertainty modeling
    • 2021.5 - 2021.11, Intern, Megvii, Few-shot learning