威廉希尔学术报告
The signal and image reconstruction using non-convex methods
葛焕敏
(北京体育大学)
报告时间:2023年11月24日 星期五 下午16:00-17:00
报告地点:沙河校区E404
报告摘要: In the report, we first consider non-convex minimization methods, which receives considerable attentions recently due to their effectiveness for signal and image recovery, and establish corresponding theoretical analysis. Then, combining the nonconvex method with deep image prior, we propose the optimization technique (DIP-AITV) for image denoising and recovery. And a nonconvex minimization approach is also proposed to solve the tensor completion problem, which is more effective to further approximate the tensor-rank.
报告人简介:葛焕敏,博士,北京体育大学副教授,硕士生导师,毕业于中国工程物理研究院获基础数学博士学位。从事应用调和分析,逼近论以及运动表现分析相关领域的研究。研究内容为信号与图像处理,运动表现分析的数学问题与方法。主持国家自然科学基金2项,北京市自然科学基金面上项目1项,主持科技部国家重点研发计划项目子课题1项。在Science China-Mathematics、IEEE Transactions on Information Theory、IEEE Signal Processing Letters、Inverse Problems 和 SIAM J. Imaging Sciences等期刊发表论文10余篇。
邀请人: 谢家新