Graph Machine Learning for Visual Computing

CVPR 2022 Tutorial

Time:   June 20th

Hybrid Event

Great Hall B

 CVPR Virtual Site

Graph Machine Learning for Visual Computing Tutorial @ CVPR 2022

Abstract: Advances in convolutional neural networks and recurrent neural networks have led to significant improvements in learning on regular grid data domains such as images and text. However, many real-world datasets or underlying relations do not lie in such simple grid structures. Such data is irregular or non-Euclidean in structure and has complex relational information. Graph Machine Learning, especially Graph Neural Networks (GNNs), provides a potential solution for processing such irregular data and for modeling the relation between entities. Numerous data formats in the visual computing area such as point clouds, 3D meshes, scene graphs, etc. have such complex structures making it challenging to model their representation for machine learning tasks. Graph Machine Learning presents powerful tools to approach this representation learning problem. Recent works have successfully applied Graph Machine Learning to handle various problems in the visual computing area such as geometric processing, scene graph generation, video understanding, multi-object relational mining, physical reasoning from vision, graphics simulation, visual navigation and so on. To spark more interests and creative thoughts, we believe it is important and beneficial to further spread the knowledge of Graph Machine Learning to the computer vision community. This tutorial will cover a wide variety of topics such as the core theory of graph machine learning, its applications in visual computing, and an introduction to one of the most popular graph ML programming frameworks.


Time Session Speakers Recordings
{{tableData[currentCountry][0]}} Opening Remarks Organizers video
{{tableData[currentCountry][1]}} Geometric Deep Learning Petar Veličković video
{{tableData[currentCountry][2]}} Building GNNs with Pytorch Geometric Matthias Fey video
{{tableData[currentCountry][3]}} Graph ML for Video Understanding Bernard Ghanem video
{{tableData[currentCountry][4]}} 20 Minutes Break
{{tableData[currentCountry][5]}} Scene Graphs in 3D Vision Federico Tombari& Fabian Manhardt video
{{tableData[currentCountry][6]}} Physion: Evaluating physical prediction from vision in humans and machines Judith Fan video
{{tableData[currentCountry][7]}} Hierarchical 3D Scene Understanding for Robotics: Real-time Algorithms and Performance Guarantees Luca Carlone & Rajat Talak video
CVPR 2022 Tutorial ©2022