DAR: Multi-View Multi-Scale Driver Action Recognition with Vision Transformer.
Yunsheng MaLiangqi YuanAmr AbdelraoufKyungtae HanRohit GuptaZihao LiZiran WangPublished in: CVPR Workshops (2023)
Keyphrases
- multi view
- action recognition
- multiscale
- computer vision
- view invariant
- single view
- human actions
- multiple views
- depth map
- activity recognition
- three dimensional
- camera calibration
- multiple cameras
- d objects
- image processing
- vision system
- body parts
- view synthesis
- human activities
- motion capture
- semi supervised
- image representation
- range images
- image sequences
- multiple viewpoints
- keypoints
- object detection
- semi supervised learning
- image classification