Binbin Huang

PhD Student at The University of Hong Kong

I work on 3D computer vision. My goal is to give machines a human-like sense of the physical world. This requires solving two problems: how to recover what a camera can see, and how to imagine what it cannot. My research therefore spans 3D reconstruction, differentiable rendering, and generative modeling. I am advised by Shenghua Gao and collaborate with Yi Ma. I have spent wonderful years at ShanghaiTech, MSRA, and SCUT.

Recent Research

See Google Scholar for a complete list of publications. · * indicates equal contribution.

	CUPID: Generative 3D Reconstruction via Joint Object and Pose Modeling Binbin Huang, Haobin Duan, Yiqun Zhao, Zibo Zhao, Yi Ma, Shenghua Gao CVPR, 2026 (Highlight) project page / arXiv / Code Create canonically posed 3D object and an object-centric camera from single image in just a few seconds.
	GenFusion: Closing the Loop between Reconstruction and Generation via Videos Sibo Wu, Congrong Xu, Binbin Huang, Andreas Geiger, Anpei Chen CVPR, 2025 project page / arXiv / Code Generative scene inpainting with a reconstruction-driven video generation model.
	2D Gaussian Splatting for Geometrically Accurate Radiance Fields Binbin Huang, Zehao Yu, Anpei Chen, Andreas Geiger, Shenghua Gao SIGGRAPH, 2024 (Most Influential Paper#1) project page / arXiv / Code Surfels and a ray-cast based differentiable rasterizer enable efficient and high-fidelity geometry and appearance reconstruction from real images.
	Mip-Splatting: Alias-free 3D Gaussian Splatting Zehao Yu, Anpei Chen, Binbin Huang, Torsten Sattler, Andreas Geiger CVPR, 2024 (Oral, Best Student Paper) project page / arXiv / Code Mip-filters enables synthesizing alias-free scenes with 3D Gaussian Splatting.
	3D-aware Image Generation using 2D Diffusion Models Jianfeng Xiang, Jiaolong Yang, Binbin Huang, Xin Tong ICCV, 2023 project page / arXiv / Code Create 3D consistent scene from a single image via iterative multiview RGBD sampling and fusion.
	Omni-Line-of-Sight Imaging for Holistic Shape Reconstruction Binbin Huang, Xingyue Peng, Siyuan Shen, Suan Xia, Ruiqian Li, Yanhua Yu, Yuehan Wang, Shenghua Gao, Wenzheng Chen, Shiying Li, Jingyi Yu arXiv preprint, 2023 arXiv Omni-line-of-sight imaging enables holistic shape reconstruction by capturing both visible and occluded geometry.
	Look Before You Leap: Learning Landmark Features for One-Stage Visual Grounding Binbin Huang, Dongze Lian, Weixin Luo, Shenghua Gao CVPR, 2021 arXiv / Code A context-aware one-stage vision-language model to localize textual queries within cluttered scenes more accurately.

Service

Journal reviewer: TPAMI, TVCG ...

Conference reviewer: SIGGRAPH, SIGGRAPH Asia, CVPR, ICCV, ECCV, ICLR, NeurIPS ...