GCHGAT: pedestrian trajectory prediction using group constrained hierarchical graph attention networks.
Lei ZhouYingli ZhaoDingye YangJingtai LiuPublished in: Appl. Intell. (2022)
Keyphrases
- prediction accuracy
- average degree
- prediction algorithm
- edge weights
- fully connected
- small world
- structured data
- prediction error
- dynamic networks
- path length
- graph representation
- graph theory
- random walk
- graph matching
- prediction model
- object detection
- network size
- graph structures
- protein function prediction
- weighted graph
- graph theoretic
- social graphs
- directed graph
- protein interaction networks
- graph structure
- graph layout
- community discovery
- social networks
- directed edges
- densely connected
- trajectory data
- biological networks
- graph databases
- network analysis
- pedestrian detection
- visual attention
- network structure