SGCN: Sparse Graph Convolution Network for Pedestrian Trajectory Prediction.
Liushuai ShiLe WangChengjiang LongSanping ZhouMo ZhouZhenxing NiuGang HuaPublished in: CVPR (2021)
Keyphrases
- densely connected
- prediction accuracy
- small world
- graph representation
- random walk
- network structure
- network traffic
- graphical representation
- computer networks
- network size
- gaussian graphical models
- wireless sensor networks
- protein interaction networks
- weighted graph
- radial basis function network
- random graphs
- strongly connected
- power law
- fully connected
- spanning tree
- prediction error
- graph theory
- pedestrian detection
- bipartite graph
- prediction model
- complex networks
- high dimensional
- neural network
- location prediction
- elman network
- community discovery
- functional modules
- trajectory data
- path length
- graph theoretic
- directed acyclic graph
- regression model
- image processing