Title
Texture Filtering in Deep Learning Era
Abstract
Texture filtering is a fundamental problem in computational photography and image analysis. It aims to smooth out textures while preserving the prominent image structures. This task receives considerable research attention, since it not only benefits scene understanding but also enables a wide variety of image manipulation applications. Most previous methods in this area are not deep learning based, due to the difficulty of training data collection. In this talk, I will share our latest research progress on texture filtering, where an effective traditional method built upon image Pyramid along with a novel self-supervised texture filtering approach will be introduced.
Biographies
Dr. Qing Zhang is now an Associate Professor in the School of Computer Science and Engineering at the Sun Yat-sen University. His main research interests are in the joint field of computer graphics and computer vision, especially computational photography and 3D vision. His research works have been published in top-tier journals and conferences, e.g., SIGGRAPH/CVPR/NeurIPS/ACM TOG/IEEE TPAMI. He won the 2022 World Artificial Intelligence Conference Youth Outstanding Paper Award.
