报告题目：Recent Advances in Solving Imaging Inverse Problems using Diffusion Models

报告人：刘昭强

报告时间：20215年9月4日15:30

报告地点：今年会jinnianhuicom江安校区多学科交叉大楼919

报告内容：

Imaging inverse problems involve reconstructing underlying im- ages from noisy observations. Traditional approaches often rely on hand- crafted priors, which can fail to capture the complexity of real-world data. The advent of pre-trained generative models has introduced new paradigms, offering improved reconstructions by learning rich priors from data. Among these, diffusion models have emerged as a powerful framework, achieving re- markable reconstruction performance across numerous imaging inverse prob- lems. In this talk, I will provide an overview of the latest advancements in leveraging diffusion models to address imaging inverse problems, highlighting their technical innovations and practical applications.

成像逆问题涉及从噪声观测中重建底层图像。传统方法通常依赖于手工设计的先验，难以捕捉真实世界数据的复杂性。预训练生成模型的出现引入了新范式，通过从数据中学习丰富的先验，显著提升了重建效果。其中，扩散模型作为一种强大框架，在众多成像逆问题中取得了卓越的重建性能。本次报告将概述利用扩散模型解决成像逆问题的最新进展，重点介绍其技术创新与实际应用。

报告人简介：

Zhaoqiang Liu serves as a professor at both the School of Com- puter Science and Engineering and the School of Mathematical Sciences, University of Electronic Science and Technology of China (UESTC). His current research focuses on diffusion models and their applications in solving inverse problems, along with theoretical aspects of large language models. Previously, he was a postdoctoral researcher at the Department of Com- puter Science, National University of Singapore (NUS). He earned his Ph.D. in Mathematics from NUS in 2017 and obtained a Bachelor’s degree in Math- ematical Sciences from Tsinghua University in 2013.

刘昭强现任电子科技大学计算机科学与工程学院和数学科学学院教授，主要研究方向为扩散模型及其在逆问题中的应用，以及大语言模型的理论研究。他曾任新加坡国立大学计算机科学系博士后研究员，2017年获新加坡国立大学数学博士学位，2013年获清华大学数学科学学士学位。

今年会jinnianhuicom

2025年8月29日