Free-Form Composition Networks for Egocentric Action Recognition.
Haoran WangQinghua ChengBaosheng YuYibing ZhanDapeng TaoLiang DingHaibin LingPublished in: CoRR (2023)
Keyphrases
- action recognition
- free form
- activity recognition
- human actions
- bag of words
- action classification
- body parts
- computer vision
- recognition of human actions
- recognizing actions
- human detection
- spatial temporal
- complex scenes
- human activity recognition
- human activities
- recognizing human actions
- bag of features
- action primitives
- depth sensors
- free form surfaces
- view invariant action recognition
- static images
- human pose
- video dataset
- max margin
- multi view
- vision system
- object detection
- spatio temporal
- viewpoint
- image processing