Home

I am a professor in the School of Computer Science and Engineering and School of Mathematical Sciences at the University of Electronic Science and Technology of China (UESTC). Currently, I mainly work on diffusion models and their applications for solving inverse problems, as well as the theoretical aspects of large language models. Previously, I was a postdoc with Prof. Jonathan Scarlett at the Department of Computer Science, National University of Singapore (NUS). I received my Ph.D. degree from the Department of Mathematics at NUS in 2017, under the supervision of Prof. Weizhu Bao and Prof. Vincent Yan Fu Tan. I obtained my Bachelor's degree from the Department of Mathematical Sciences of Tsinghua University in 2013.

Always looking for self-motivated Ph.D./master students with strong mathematical/computer-science background.

Journal reviewing note: Please get in touch with me personally before sending review invitations. I receive many unsolicited review invitations, which are automatically identified as spam and deleted before I can read them.

News

(Sep. 2026) Paper Recovery guarantees for posterior sampling of one-bit compressed sensing accepted to NeurIPS 2026

(Aug. 2026) Paper A posterior-dynamics framework for imaging inverse problems with pretrained diffusion priors uploaded to arXiv [Paper] 

(Jun. 2026) Paper Wavelet-guided semantic signal compensation for inversion-free image editing accepted to ECCV 2026 [Paper] [Code]

(May 2026) Paper Tracing the oracle: Improving diffusion timestep scheduling for 3D CT reconstruction accepted to ECML-PKDD 2026 [Paper] [Code]

(May 2026) Recognized as a Silver Reviewer for ICML 2026

(May 2026) Paper Image restoration via diffusion models with dynamic resolution accepted to ICML 2026 [Paper] [Code]

(Feb. 2026) Papers Just-in-Time: Training-free spatial acceleration for diffusion Transformers [Paper] [Project Page] and Outlier-robust diffusion solvers for inverse problems [Paper] [Code] accepted to CVPR 2026 

(Jan. 2026) Paper Operator splitting based diffusion samplers and improved convergence analysis uploaded to arXiv [Paper]

(Jan. 2026) Six papers accepted to ICASSP 2026 [Papers]

(Nov. 2025) Paper Diffusion model based signal recovery under 1-bit quantization accepted to AAAI 2026 [Paper] [Code]

(Sep. 2025) Paper Learnable sampler distillation for discrete diffusion models accepted to NeurIPS 2025 (Spotlight Presentation) [Paper] [Code]

(May 2025) Papers Integrating intermediate layer optimization and projected gradient descent for solving inverse problems with diffusion models [Paper] [Code] and Learning single index models with diffusion priors [Paper] accepted to ICML 2025 

(Jan. 2025) Paper AlgoFormer: An efficient Transformer framework with algorithmic structures accepted to Transactions on Machine Learning Research [Paper]

(Jan. 2025) Paper Solving quadratic systems with full-rank matrices using sparse or generative priors accepted to IEEE Transactions on Signal Processing [Paper]

(Sep. 2024) Paper Generalized eigenvalue problems with generative priors accepted to NeurIPS 2024 [Paper]

(Feb. 2024) Paper Accelerating diffusion sampling with optimized time steps accepted to CVPR 2024 [Paper] [Code]

(Jan. 2024) I have joined the School of Computer Science and Engineering, University of Electronic Science and Technology of China

Everything should be made as simple as possible, but not simpler. Albert Einstein