AST-GNN: An attention-based spatio-temporal graph neural network for Interaction-aware pedestrian trajectory prediction.
Hao ZhouDongchun RenHuaxia XiaMingyu FanXu YangHai HuangPublished in: Neurocomputing (2021)
Keyphrases
- spatio temporal
- neural network
- prediction model
- trajectories of moving objects
- prediction accuracy
- protein interaction
- human computer interaction
- space time
- object trajectories
- random walk
- moving object databases
- artificial neural networks
- movement patterns
- spatial and temporal
- multi layer perceptron
- spatio temporal data
- neural network model
- predictive model
- visual attention
- moving objects
- structured data
- neural nets
- image sequences
- location prediction
- user interaction
- prediction algorithm
- graph theory
- graph representation
- graph model
- spatio temporal databases
- protein function prediction
- trajectory data
- elman network
- computer vision
- pattern recognition
- directed acyclic graph
- prediction error
- graph structure
- bp neural network
- human actions
- self organizing maps
- back propagation
- genetic algorithm
- graph databases
- mobile robot
- fuzzy logic
- graph matching
- graph mining
- fuzzy neural network
- protein protein interactions