| CARVIEW |
About me
I'm Zhiyuan Ma, an undergraduate of Peking University, class of 2022. I major in computer science, and is selected for the Top Selection plan(bajian plan). Now I'm studying embodied AI in professor Hao Dong 's PKU-Agibot lab, following the guidance of Jiyao Zhang.
age : 20
address : Peking University
GPA : 3.605
hobbies : guitar,singing,ping pong
email : 1829142919@qq.com
CET-6 : 564
Experience
Undergraduate at Peking University
Selected for Bajian Plan
PKU-EPIC & Galbot joint lab
PKU-Hyperplane & Agibot Lab
Intern at AIoT Group of Prof. Xinlei Chen, Tsinghua SIGS
Research
Intern at PKU-EPIC Lab (Mentor: Jiayi Chen)
I studied deep learning, 3D representations, and fundamental robotics. I assisted in organizing articulated object datasets (GAPartNet, UnidoorManip, AKB48) and implemented some code practices based on DexGraspNet and curobo. I presented papers in group meetings and participated in lab discussions every week.
July.2024 - nowCross-Embodiment Dexterous Grasping in Cluttered Scenes
This is my current ongoing project, mentored by Jiyao Zhang and Tianhao Wu. We employ a hand-agnostic grasp pose representation method, namely the improved IBS (Interaction Bisector Surface), to guide our model in grasping with unseen robotic hands in cluttered scenes. We designed a two-stage generative network to predict IBS representations from scene observation point clouds, which is used to guide hand pose optimization, and finally execute scene-level grasping on the physical robot.
Publications
CADGrasp: Learning Contact and Collision Aware General Dexterous Grasping in Cluttered Scenes
Jiyao Zhang*, Zhiyuan Ma*, Tianhao Wu, Zeyuan Chen, Hao Dong
Accepted by NeurIPS 2025We propose CADGrasp, which learns a contact- and collision-aware inter- mediate representation as a constraint, and further obtains the dexterous grasp pose with an optimization method to achieve single-view dexterous hand grasping in cluttered scenes. Our method can easily transferred to zero-shot robotic hands, showing cross-embodiment generalization ability. Accepted by NeurIPS 2025.
Awards
Merit Student
in the academic year of 2022-2023
Lingjun Pioneer Scholarship
in the academic year of 2022-2023
Tianchuang Scholarship
in the academic year of 2023-2024
Xiaomi Scholarship
in the academic year of 2024-2025
School-level Merit Student
in the academic year of 2024-2025
contact info
my address
Peking University
Beijing Haidian District
my number
(+86)18810685032
my email
1829142919@qq.com
2200013163@stu.pku.edu.cn