Self-Growing Spatial Graph Networks for Pedestrian Trajectory Prediction.
Sirin HaddadSiew-Kei LamPublished in: WACV (2020)
Keyphrases
- prediction error
- prediction accuracy
- spatio temporal
- average degree
- fully connected
- highly connected
- graph representation
- spatial data
- spatial information
- dynamic networks
- prediction model
- structured data
- graph theory
- social networks
- edge weights
- complex networks
- network structure
- spatial and temporal
- space time
- random walk
- inter layer
- community discovery
- neural network ensemble
- prediction algorithm
- link formation
- graph model
- connected components
- network analysis
- weighted graph
- spatial relationships
- graph structure
- neural network
- graph mining
- social graphs
- graph matching
- location prediction
- citation networks
- protein interaction networks
- small world
- network size
- pedestrian detection
- graph structures
- spatial relations
- spatial networks
- path length
- overlapping communities
- biological networks
- community structure
- community detection
- densely connected
- graph layout
- continuously moving
- graph databases