|
Chen Si | 司宸 Starting in August 2026, I will be an incoming first-year Ph.D. student in the College of Computing and Data Science (CCDS) at Nanyang Technological University (NTU), supervised by Prof. Chuanxia Zheng in the Physical Vision Group (PVG). Prior to that, I received my M.S. in Computer Science from University of California San Diego (UCSD) and my B.S. in Computer Science from the University of Illinois Urbana-Champaign (UIUC). During that time, I've been fortunate to collaborate with Prof. Hao Su, Prof. Romit Choudhury, and Prof. Lianhui Qin. I am interested in multimodal AI, with a current focus on computer vision and robotics. My research explores how models can robustly perceive, reconstruct, and interact with the physical world using signals such as RGB, depth, geometry, and dynamics. I am particularly interested in physically grounded perception for embodied systems, especially in 3D vision and hand-object interaction, with the goal of supporting simulation and real-world robotic applications. / CV / GitHub / Google Scholar / LinkedIn |
|
News
|
ResearchMy research focuses on multimodal AI, especially computer vision and robotics. I study how models can perceive, reconstruct, and interact with the physical world using signals such as RGB, depth, geometry, and dynamics. I am particularly interested in 3D vision, hand-object interaction, and physically grounded perception for embodied systems. Publications are listed by recency. Key papers are in the Selected tab. |
|
AnyHand: A Large-Scale Synthetic Dataset for RGB(-D) Hand Pose Estimation
, , , , , , Website / arXiv / code
TL;DR:
AnyHand is a large-scale synthetic dataset enabling robust RGB(-D) hand pose estimation across diverse scenarios.
|
|
ArcMemo: Abstract Reasoning Composition with Lifelong LLM Memory
, , , , , , arXiv
TL;DR:
ArcMemo is a memory-augmented agent framework for ARC that improves puzzle solving by storing, retrieving, and reusing useful intermediate reasoning patterns across tasks.
|
|
Explicit Context-Driven Neural Acoustic Modeling for High-Fidelity RIR Generation
Interspeech 2026
, , , Website / arXiv / code
TL;DR:
MiNAF leverages explicit local geometric information from 3D room meshes for fast, accurate, and high-fidelity room impulse response prediction.
|
|
Coupled management of electric vehicle workplace charging and office building loads
, , , Paper
Transportation Research Part D: Transport and Environment, Volume 134, September 2024
TL;DR:
A framework that jointly manages workplace EV charging and office building energy use, showing that smart building controls and optimized charging schedules can cut costs across different U.S. climates.
|
Selected Honors
|
Experience |
|
Nanyang Technological University, Singapore 2026.08 - PresentPh.D. Student, College of Computing and Data Science (CCDS) |
|
University of California San Diego, USA 2024.09 - 2026.06M.S. in Computer Science |
|
University of Illinois Urbana-Champaign, USA 2020.08 - 2024.05B.S. in Computer Science |
|
Thank you for visiting! Feel free to contact me if you have any questions.
|