GeometryMotion-Net: A Strong Two-Stream Baseline for 3D Action Recognition.
Jiaheng LiuDong XuPublished in: IEEE Trans. Circuits Syst. Video Technol. (2021)
Keyphrases
- action recognition
- human actions
- bag of words
- computer vision
- data streams
- human detection
- spatial temporal
- action classification
- activity recognition
- body parts
- recognition of human actions
- recognizing human actions
- human pose
- static images
- depth sensors
- independent subspace analysis
- human activities
- machine learning
- object detection
- probabilistic model
- view invariant
- action recognition in videos