Graph Signal Processing

The goal of this lecture is to convey the following skills:

  • understand the fundamentals of Graph Signal (GS) representation and GS processing based on spectral graph theory
  • obtain an overview and technical depth of some methods for graph filtering and sampling
  • apply GSP methods to a range of areas including the analysis of distributed sensor networks and point clouds

This is a theory lecture with applications from volumetric multimedia processing and machine learning. The lecture is given in English.

Contents

  • Short introduction to graph signals and node domain processing
  • Node domain graph filters
  • Graph Fourier Transform, Filtering 
  • Application of GFT to common operators
  • Graph Spectra
  • Graph Signal models, node domain sampling, frequency domain sampling, Conditions for reconstruction
  • Robust Graph spectral sampling
  • Applications of GSP to domains including transportation networks, sensor networks, point clouds, and learning with Graph Signals

Dates

Summer term

Lecture

  • Mo, 9:00 - 10:30, Room 3403 - A141
  • Start Lecture: 14.04.2025

Exercise

  • Mo 10:30 - 11:50, Room 3403 -  A141
  • Start Exercise: 14.04.2025

Materials

Materials will be distributed via Stud.IP

Exam

Written Exam (120 min)

Prüfungs- und Vorlesungsanmeldung


Ihr Dozent

Prof. Dr. Amr Rizk
Professors
Address
Appelstraße 9a
30167 Hannover
Building
Room
Prof. Dr. Amr Rizk
Professors
Address
Appelstraße 9a
30167 Hannover
Building
Room