【报告主题】Nonlinear Techniques in Graph Signal Processing

【报告人】Dr David Tay, Deakin University

【报告时间】2026年4月8日（周三）下午2:30

【报告地点】北京大学王选计算机研究所106报告厅

Abstract: Graph Signal Processing (GSP) seeks to develop techniques for processing signal defined over irregular domains, that can be modelled as graphs. The area has had increasing interest over the last decade and has had successful applications in a variety of areas such as social networks, sensor networks, bioinformatics and image processing. One the main goals of GSP is to generalize Classical Signal Processing (CSP) tools for graph signals, e.g. spectral filtering and signal transforms. Most of the developments in GSP are based on linear processing, e.g. Graph-Fourier-Transform. Though not as popular as linear techniques, nonlinear techniques have had some success in CSP. One such class of techniques is based on the median operator, and in some applications, e.g. denoising impulsive noise in images, is substantially superior to the linear counterpart.

In this talk, we will describe some of our work in exploiting the median operator in GSP. We first describe several non-linear graph filters that are based on the median operator, and their application to graph signal denoising. We then describe the development of non-linear graph wavelet transforms by converting a linear transform into a non-linear equivalent, through a process termed medianfication. The resulting nonlinear transform shares many desirable properties of the corresponding linear transform, but also has the edge-aware property, whereby filtering across signal discontinuities which results in distortions, can be avoided.

Bio: David B. Tay received the B.Eng. degree in electrical and electronic engineering and the B.Sc. degree in mathematics from the University of Melbourne, Melbourne, VIC, Australia, and the Ph.D. degree in signal processing from Cambridge University, Cambridge, U.K. He was previously a Lecturer and an Assistant Professor with the School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. He was then with the Department of Engineering, LaTrobe University, first as a Lecturer and subsequently as an Associate Professor. He is currently a Senior Lecturer with the School of Information Technology, Deakin University, Melbourne, Australia. He is a member of the DSP Technical Committee in the IEEE Circuits and Systems Society, and has served as an editor in The Journal of Franklin Institute.