Referring Expression Grounding by Marginalizing Scene Graph Likelihood.
Daqing LiuHanwang ZhangZheng-Jun ZhaFanglin WangPublished in: CoRR (2019)
Keyphrases
- d scene
- video sequences
- indoor and outdoor
- structured data
- random walk
- single image
- input image
- graph theory
- scene analysis
- scene understanding
- directed graph
- complex scenes
- graph structure
- weighted graph
- graph representation
- image sequences
- graph model
- connected components
- maximum likelihood
- multiple images
- graph mining
- graph theoretic
- three dimensional
- bipartite graph
- dynamic scenes
- ground plane
- graph partitioning
- moving objects
- graph based algorithm
- scene images