About

I am a first-year Ph.D. student in Computer Science at the University of Illinois Urbana-Champaign (UIUC), co-advised by Prof. Shenlong Wang and Prof. Alexander Schwing.

Prior to this, I obtained my M.S. degree from UIUC and my B.S. degree from National Tsing Hua University (NTHU). I was grateful to be selected as a Siebel Scholar during my Master's.

My research interests lie in 3D vision, generative modeling, and large multimodal models (VLM, MLLM). I am broadly interested in building computational systems that can perceive, represent, and interact with our physical world.

News

* indicates equal contribution.

NeurIPS 2025

HoloScene: Simulation-Ready Interactive 3D Worlds from a Single Video

Hongchi Xia, Chih-Hao Lin, Hao-Yu Hsu, Quentin Leboutet, Katelyn Gao, Michael Paulitsch, Benjamin Ummenhofer, Shenlong Wang

We reconstruct simulation-ready and interactable digital twin assets from a single video.

arXiv / website / code /

ICCV 2025

PhysTwin: Physics-Informed Reconstruction and Simulation of Deformable Objects from Videos

Hanxiao Jiang, Hao-Yu Hsu, Kaifeng Zhang, Hsin-Ni Yu, Shenlong Wang, Yunzhu Li

We optimize a spring-mass physics model of deformable objects and integrate it with 3D Gaussian Splatting for real-time re-simulation with rendering.

arXiv / website / code /

3DV 2025

AutoVFX: Physically Realistic Video Editing from Natural Language Instructions

Hao-Yu Hsu, Chih-Hao Lin, Albert J. Zhai, Hongchi Xia, Shenlong Wang

A system that generates dynamic, physically realistic visual effects (VFX) on a single video solely from text-based editing instructions.

arXiv / website / code /

NeurIPS 2022

SPoVT: Semantic Prototype Variational Transformer for Dense Point Cloud Semantic Completion

Sheng-Yu Huang*, Hao-Yu Hsu*, Yu-Chiang Frank Wang

A point cloud semantic completion framework that completes partial point clouds of 3D objects with a variational Transformer.

paper / website / video / code /

CVPR 2022

NeurMiPs: Neural Mixture of Planar Experts for View Synthesis

Zhi-Hao Lin, Wei-Chiu Ma*, Hao-Yu Hsu*, Yu-Chiang Frank Wang, Shenlong Wang

Uses an efficient 3D planar representation to model the geometry and appearance of a scene for novel view synthesis.

arXiv / website / video / code /

Experience

Research Assistant

Vision & Learning Lab, National Taiwan University

Sep. 2021 – Feb. 2023

Working with Prof. Yu-Chiang Frank Wang and Prof. Shao-Hua Sun on 3D vision and robot learning.

Awards & Service

• Amazon AI PhD Fellowship, 2025–2026 [Amazon News]
• Siebel Scholar, Class of 2025 [UIUC News] [Siebel Foundation News]
• Phi Tau Phi Scholastic Honor Society Honorary Membership (Graduated with the first place in NTHU, Class of 2021)
• Reviewer: CVPR, ECCV, NeurIPS, BMVC, SIGGRAPH Asia, 3DV, AAAI, ICRA, TPAMI

Misc

• I love various forms of sport. I often play basketball 🏀 in my spare time, and also enjoy baseball ⚾, swimming 🏊, weightlifting 🏋️, and cycling 🚴.