Capsule Networks - A Survey by Dr. Yogesh Rawat, UCF

Explore the innovative concept of Capsule Networks in this comprehensive 29-minute survey presented by Dr. Yogesh Rawat from the University of Central Florida. Delve into various aspects of Capsule Networks, including their application in computer vision, as discussed in the CVPR 2019 Tutorial. Learn about generalization to video, dynamic and EM routing, generic routing by agreement, and the principles of equivariance and invariance. Discover fast dynamic routing techniques, alternative notions of capsules, and their applications in text, graphs, 3D point clouds, and multi-modal scenarios. Examine deeper capsule network architectures and their problem domains, including segmentation and localization. Investigate practical applications such as relation extraction, adversarial attack detection, reconstruction, and medical image classification for brain tumors and breast cancer. Gain insights into the challenges and open problems in the field, and access valuable resources for further study.

Intro
Outline
Introduction
Generalization to video
Dynamic vs EM routing
Generic routing by agreement
Equivariance and Invariance
Fast dynamic routing
Another notion of capsule
Text capsules
Graph capsules
3D point cloud
Multi-modal
Deeper capsule network
Problem domain
Segmentation
Localization
Applications
Relation extraction
Detecting adversarial attack
Reconstruction
Brain tumor classification
Breast cancer classification
Challenges and open problems
Resources