A motion-aware and temporal-enhanced Spatial-Temporal Graph Convolutional Network for skeleton-based human action segmentation.
Shurong ChaiRahul Kumar JainJiaqing LiuShiyu TengTomoko TateyamaYinhao LiYen-Wei ChenPublished in: Neurocomputing (2024)
Keyphrases
- spatial temporal
- human actions
- space time
- human motion
- spatial and temporal
- action recognition
- spatio temporal
- motion recognition
- motion history images
- temporal information
- video sequences
- convolutional network
- segmentation algorithm
- image segmentation
- image sequences
- temporal sequences
- motion patterns
- human body
- convolutional neural networks
- computer vision
- video shots
- motion model
- video database
- activity recognition
- human activities
- moving objects
- shape prior
- machine learning
- three dimensional
- reinforcement learning
- key frames
- background subtraction
- image registration
- visual features