Academic Report of the School of Mathematical Sciences  No. 059  

(Series Report on High-Level University Construction No. 1081)  

Lecture Title: Variational Bayesian Inference from Robust Tensor Analysis to Spatially-Variant Deblurring

Speaker: Dr. Chao Wang (Assistant Professor, Southern University of Science and Technology)  

Date & Time: July 1, 2025, 10:00–11:00 AM  

Venue: Alumni Plaza, Room 307  

Abstract: Variational Bayesian inference provides a powerful framework for addressing critical challenges in tensor analysis and image restoration—key problems in machine learning and computer vision. This lecture introduces innovative methods that integrate Bayesian principles, variational inference, and physical modeling. First, a Bayesian framework for Tensor Robust Principal Component Analysis (TRPCA) is proposed to recover low-rank structures and characterize sparse noise in mixed-noise scenarios. By embedding a low-rank tensor nuclear norm prior and a generalized sparsity-inducing prior within a Bayesian framework, the method automatically learns the optimal tensor nuclear norm and balances low-rank and sparse components. Second, a novel physics-informed optimization framework is developed for jointly solving depth estimation and image restoration from a single defocused image. By modeling the defocused image as a function of a depth map and an all-in-focus (AiF) image based on optical physics, the framework leverages their intrinsic connections. The depth map guides AiF image recovery, while the AiF image regularizes depth map reconstruction via reconstruction error. A variational inference algorithm parameterized with deep neural networks ensures flexibility and high performance.

Biography: Dr. Chao Wang is an Associate Researcher and PhD Supervisor at the Department of Statistics and Data Science, Southern University of Science and Technology. His research focuses on image processing, scientific computing, and interdisciplinary data science. He has published over 30 papers in top-tier journals and conferences, including *SIAM* series and *IEEE* transactions. Awards include the Best Paper Award at the 2022 CVPR Workshop , the Shenzhen Pengcheng Kingfisher Talent Program  (2021), and the Best Paper Award at the 2017 Annual Conference of the Chinese Society for Industrial and Applied Mathematics . He leads the National Natural Science Foundation for Young Scientists  and participates in key national R&D projects and Hong Kong Research Grants Council-funded initiatives as principal investigator or core member.

All faculty and students are welcome!

Invited by:  Wang Jiangzhou

School of Mathematical Sciences

June 26, 2025