Welcome!

I am Xiaoyu Zhang (张啸宇), a Research Scientist at XPENG Robotics, where we are working on scalable robotic foundation models. I received my Ph.D. from The Chinese University of Hong Kong (CUHK) under the supervision of Prof. Yun-Hui Liu. I earned my M.S. and B.S. degrees from Beihang University (BUAA), advised by Prof. Wei Wang. I have also had the privilege of collaborating with Dr. Ji Zhao and Prof. Shiyu Zhao.

My research interests now center on vision-language-action (VLA) models and foundational models for robotics. Previously, I also worked on computer vision for autonomous driving, SLAM, and robotic perception.

🔥 News

• 2026.05:  🎉🎉 We release Fe₀, exploring what embodied foundation model learns from cross-embodiment data.
• 2026.01:  🎉🎉 I join XPENG Robotics as a Research Scientist!
• 2026.01:  🎉🎉 I have passed my Ph.D. defense!
• 2024.08:  🎉🎉 One paper on memory-efficient SLAM is accepted by JFR!
• 2024.07:  🎉🎉 Two papers on HD map construction are accepted by ECCV 2024!
• 2023.02:  🎉🎉 One paper on map sparsification is accepted by CVPR 2023!

📝 Recent Publications

For the full lists, please refer to my Google Scholar page

*Equal contribution

XPENG Robotics Blog

Inside Fe₀: What Cross-Embodiment Data Teaches an Embodied Foundation Model

Yizhuo Li^*, Xiaoyu Zhang^*, Yuying Ge^*#, Teng Wang, Boyu Chen, Yi Chen,
Bo Liu, Feng Qiu, Yuguo Gan, Hui Zhou, Yixiao Ge

XPENG Robotics Blog, 2026

[blog]

JFR 2024

MS-SLAM: Memory‐Efficient Visual SLAM with Sliding Window Map Sparsification

Xiaoyu Zhang, Jinhu Dong, Yin Zhang, and Yun-hui Liu

Journal of Field Robotics (JFR), 2024

[paper] [code]

ECCV 2024

Enhancing Vectorized Map Perception with Historical Rasterized Maps

Xiaoyu Zhang^*, Guangwei Liu^*, Zihao Liu, Ningyi Xu, Yun-hui Liu, and Ji Zhao

The European Conference on Computer Vision (ECCV) 2024

[paper] [code]

ECCV 2024

Leveraging Enhanced Queries of Point Sets for Vectorized Map Construction

Zihao Liu^*, Xiaoyu Zhang^*, Guangwei Liu^*, Ji Zhao, and Ningyi Xu

The European Conference on Computer Vision (ECCV) 2024

[paper] [code]

CVPR 2023

Efficient Map Sparsification Based on 2D and 3D Discretized Grids

Xiaoyu Zhang, and Yun-hui Liu

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023

[paper] [code]

📖 Educations

• 2021 - 2026, Ph.D. in Mechanical and Automation Engineering, The Chinese University of Hong Kong
  • T-Stone Robotics Institue, advised by Prof.Yun-hui Liu
  • Fields: 3D computer vision, self-driving
• 2017 - 2020, M.Phil. in Mechanical Engineering, Beihang University
  • Robotics Institute, advised by Prof.Wei Wang
  • Fields: SLAM, mobile robotics
• 2013 - 2017, B.S. in Mechanical Engineering, Beihang University

💻 Work Experience

• 2020 - 2021, Westlake University
  • Research assistant, advised by Prof. Shiyu Zhao
  • Fields: computer vision for UAV
• 2019, DiDi Chuxing
  • Intern in SLAM, The Group of Visual Computing.

🎖 Honors and Awards

• 2021-2025, CUHK Vice-Chancellor’s PhD Scholarship
• 2017, First Prize in ``Challenge Cup” Capital College Student Curricular Academic Science and Technology Works Competition
• 2017, Beijing Outstanding Graduate
• 2015, National Scholarship
• 2014, National Scholarship