Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and Graph Attention Networks.
Vineet KosarajuAmir SadeghianRoberto Martín-MartínIan D. ReidSeyed Hamid RezatofighiSilvio SavaresePublished in: CoRR (2019)
Keyphrases
- social graphs
- social networks
- link formation
- small world
- social network analysis
- average degree
- short term
- random graph models
- information transfer
- complex networks
- heterogeneous social networks
- social media
- social interaction
- edge weights
- information diffusion
- network structure
- community discovery
- graph structure
- multi modal
- criminal networks
- random walk
- social graph
- overlapping communities
- graph representation
- long term
- directed edges
- fully connected
- path length
- social relationships
- prediction model
- link prediction
- graph theory
- social structure
- neural network
- connected components
- network dynamics
- power law
- network analysis
- real world networks
- dynamic networks
- weighted graph
- citation networks
- forecasting model
- directed acyclic graph
- graph mining
- graph model
- betweenness centrality
- online social networks
- community structure
- visual attention
- directed graph
- social networking sites
- graphical representation