About

I am Xiang Zhang, an AI Research Scientist at Meta working on video generation. I received my Ph.D. from UC San Diego in 2026, advised by Prof. Zhuowen Tu. Previously, I obtained my bachelor's degree in Computer Science and Technology at Tsinghua University in 2021.

My Ph.D. research focused on 3D scene reconstruction (especially from a single image), mesh generation, and controllable image generation.

News

• 2026.09Two papers accepted to NeurIPS 2026.
• 2026.08Joined Meta.
• 2026.06Received my Ph.D. from UC San Diego.
• 2026.02PixARMesh accepted to CVPR 2026.
• 2025.09OverLayBench accepted to NeurIPS 2025.
• 2025.06Four papers accepted to ICCV 2025: VertexRegen, DepR, YOLO-Count, and Lay-Your-Scene.

Selected publications

PixARMesh: Autoregressive Mesh-Native Single-View Scene Reconstruction

[pdf] [project page]

Xiang Zhang*, Sohyun Yoo*, Hongrui Wu*, Chuan Li, Jianwen Xie, Zhuowen Tu

CVPR 2026

OverLayBench: A Benchmark for Layout-to-Image Generation with Dense Overlaps

[pdf] [project page]

Bingnan Li*, Chen-Yu Wang*, Haiyang Xu*, Xiang Zhang, Ethan J. Armand, Divyansh Srivastava, Xiaojun Shan, Zeyuan Chen, Jianwen Xie, Zhuowen Tu

NeurIPS 2025

VertexRegen: Mesh Generation with Continuous Level of Detail

[pdf] [project page]

Xiang Zhang, Yawar Siddiqui, Armen Avetisyan, Chris Xie, Jakob Engel, Henry Howard-Jenkins

ICCV 2025

DepR: Depth Guided Single-view Scene Reconstruction with Instance-level Diffusion

[pdf] [project page]

Qingcheng Zhao*, Xiang Zhang*, Haiyang Xu, Zeyuan Chen, Jianwen Xie, Yuan Gao, Zhuowen Tu

ICCV 2025

Full publication list →