Shaobang Li

Title: Shaobang Li: Advancing Super-Resolution Image Reconstruction

Introduction

Shaobang Li is an inventor associated with Hangzhou, China, and Zhejiang Gongshang University. His patent work focuses on computational imaging, particularly methods for reconstructing high-resolution images from lower-resolution inputs.

Latest Patents

The provided information lists 1 patent for Shaobang Li: “Methods and Systems for Wavelet Domain-Based Normalizing Flow Super-Resolution Image Reconstruction.” The invention describes a method and system that build and train a normalizing flow model for super-resolution reconstruction. Its architecture uses multiple levels with squeeze, conditional mapping, split, activation standard, and QR layers. The method further applies a wavelet transform, a reconstructed QR layer, and a T-distribution to establish a stable model, then adds a refinement layer before training and reconstructing a super-resolution image.

Career Highlights

Li’s listed patent reflects work at the intersection of wavelet-domain processing and normalizing flow models. The approach aims to improve the stability and refinement of image reconstruction through a structured model-training process.

Collaborations

The provided information identifies Bailin Yang and Chao Song as Li’s coworkers. Their names are associated with his professional context, though specific roles in the patent are not stated.

Conclusion

Shaobang Li’s listed invention applies normalizing flows and wavelet-domain techniques to super-resolution image reconstruction, highlighting his contribution to computational imaging research.