Human Action Recognition based on ST-GCN using Opticalflow and Image Gradient.
Han-Byul JangDae-Jin KimChil-Woo LeePublished in: SMA (2020)
Keyphrases
- action recognition
- image gradient
- static images
- human activities
- human actions
- activity recognition
- computer vision
- motion history images
- action classification
- body parts
- object boundaries
- optical flow
- image structure
- recognition of human actions
- human object interactions
- recognizing actions
- view invariant
- action recognition in videos
- input image
- low level
- energy functional
- image regions
- multi view
- recognizing human actions
- viewpoint
- multiscale
- high level
- image processing