Frame-based methods together with optimization modeling have been shown to be one of the most effective approaches in imaging processing. This minisymposium focuses on solving various image processing problems, e.g., image denoising/inpainting, image restoration, pMRI and CT in medical imaging, etc., based on multiscale representation systems and optimization techniques. We will have speakers coming from mainland China, Hong Kong SAR, Canada, and USA to present their work on mathematical imaging using framelets, affine shear frames, optimization (convex or non-convex) techniques, deep neural networks, or kernel-based approaches. We believe that the audience of this minisymposium would greatly benefit from their perspectives on this area.

Moreau Enhanced TV for Image Restoration

Lixin Shen (Syracuse University)

Multiplicative Noise Removal with A Non-Convex Optimization Model

Jian Lu (Shenzhen University)

Noncovex Frame-based Methods for Image Restoration

Yi Shen (Zhejiang Sci-Tech Univeristy)

Parallel Magnetic Resonance Imaging by 3-D Regularization

Yan-Ran Li (Shenzhen Univeristy)

The Convex Geometry of Learning Single-hidden-layer Neural Networks

Gongguo Tang (Colorado School of Mines)

Digital Affine Shear Filter Banks with 2-Layer Structure and Their Applications in Image/Video Processing

Zhihua Che (City University of Hong Kong)

Framelets on Graphs with Applications in Multiscale Data Analysis

Xiaosheng Zhuang (City University of Hong Kong)

Reduced Complex Dynamical System Models and Applications to Filtering

Wonjung Lee (City University of Hong Kong)

Medical Image Analysis and Its Applications

Yao Lu (Sun Yat-sen University)

Organizers:

Bin Han (University of Alberta)

Yan-Ran Li (Shenzhen Univeristy)

Lixin Shen (Syracuse University)

Xiaosheng Zhuang (City University of Hong Kong)

Keywords:

computed tomography, deep learning, directional framelets, image deblurring, image reconstruction, image representation, inverse problems, kernel-based approximation, machine learning, manifolds and graphs