Ph.D. Student in Computer Science
University of Maryland, College Park
I am a Ph.D. student in Computer Science at the University of Maryland, College Park, advised by Prof. Ruohan Gao. I also work closely with Prof. Ming C. Lin.
My research lies at the intersection of multimodal learning, generative modeling, and 3D vision. My current work focuses on reconstructing and generating audio-visual environments from minimal observations.
Previously, I earned my B.Eng. at CUHK-Shenzhen in 2024, advised by Prof. Xiaoguang Han.
I am seeking research internship opportunities for Summer 2027.
* Equal contribution
arXiv 2026
A first framework that turns a single monocular video into an explorable 4D audio-visual scene, with spatial audio for each moving sound source kept consistent through dual-space diffusion synchronization.
Project page Paper arXiv Code (coming soon) Interactive demo
ECCV 2026
Reconstructing object-level sound fields from multi-view images and a few impact recordings for impact sound synthesis, contact localization, and sound editing.
CVPR 2026
A first generative framework for creating explorable 3D audio-visual scenes from one image, with spatial audio grounded in scene semantics and 3D geometry.
ICCV 2025, Oral Presentation
A first physics-based, audio-visual differentiable acoustic renderer that uses multi-view images and beam tracing to reconstruct room impulse responses (RIRs) efficiently and accurately from sparse real-world measurements.
NeurIPS 2023, Dataset and Benchmark Track
A unified benchmark for human NeRF models, providing standardized datasets, metrics, and a generalizable baseline.
ICCV 2023
A 3D generative framework that produces realistic and diverse digital humans with both texture and geometry by combining both 2D and 3D priors with minimal 3D supervision.
Interactive demos include spatial audio and are best experienced on a desktop with headphones.
