GAST-Net: Graph Attention Spatio-temporal Convolutional Networks for 3D Human Pose Estimation in Video.
Junfa LiuYisheng GuangJuan RojasPublished in: CoRR (2020)
Keyphrases
- spatio temporal
- spatial and temporal
- spatial temporal
- space time
- human actions
- video representation
- average degree
- video sequences
- spatio temporally
- image sequences
- video streams
- moving objects
- video data
- action detection
- multimedia
- temporal domain
- spatio temporal data
- fully connected
- weighted graph
- edge weights
- graph layout
- video database
- video frames
- graph structure
- small world
- body parts
- visual attention
- bipartite graph
- complex networks
- video analysis
- temporal structure
- graph model
- graph representation
- directed graph
- temporal segmentation
- citation networks
- random walk
- graph structures
- real time
- motion trajectories
- graph theory
- video clips
- video content
- human activities
- graph mining
- directed acyclic graph
- network analysis
- video retrieval
- degree distribution
- structured data
- temporal information
- connected components