Hi! PARIS is pleased to welcome Antonio Ortega, Professor of Electrical and Computer Engineering at the University of Southern California, for a scientific seminar organized in the framework of the Hi! PARIS International Visiting Chairs Program.

His talk, titled “How to Design Fast GFTs,” will provide an overview of recent advances in speeding up the computation of the Graph Fourier Transform, also known as GFT.

Graph Fourier Transforms play an important role in graph signal processing, with applications in image and video coding, graph machine learning, multimedia compression, 3D point cloud processing, and sensor networks. However, computing these transforms efficiently remains a key challenge, especially for large or complex graph structures.

Antonio Ortega will present divide-and-conquer techniques that make use of graph structure, including graph symmetries and graph decompositions based on low-rank updates. He will also discuss approximation methods for cases where the graph structure alone does not provide enough acceleration. These include direct transform approximations using Givens rotations, as well as indirect methods that rely on more favorable graph structures, such as spectral sparsification.

The seminar will highlight how these approaches can improve the efficiency of graph-based signal processing methods and support applications in image and video coding, graph machine learning, and related fields.

About the speaker

Antonio Ortega is Professor of Electrical and Computer Engineering at the University of Southern California. He is a Fellow of the IEEE and EURASIP, and currently serves as Vice President of Publications of the IEEE Signal Processing Society.

His research focuses on graph signal processing, including sampling, reconstruction, transforms, learning, and compression, as well as multimedia and 3D point cloud compression, distributed and error-tolerant compression, and information representation in sensor networks.

He has supervised nearly 50 Ph.D. students, authored more than 400 publications, and published the book Introduction to Graph Signal Processing with Cambridge University Press in 2022. His work has received several distinctions, including the IEEE Signal Processing Magazine Award, the ICIP Best Paper Award, and the IEEE Communications Society Leonard G. Abraham Prize.

Register here!