VTuckeR: Multimodal Tucker Fusion for Scene Graph Generation.
Geng JiaYunhui YiWentao KanPublished in: icWCSN (2022)
Keyphrases
- d scene
- random walk
- single image
- scene understanding
- complex scenes
- graph theory
- multiple objects
- three dimensional
- scene classification
- video sequences
- graph representation
- multi modal
- real scenes
- weighted graph
- graph structure
- dynamic scenes
- graph mining
- graph databases
- image sequences
- graph theoretic
- outdoor scenes
- multiple images
- directed graph
- connected components
- multimodal fusion
- image regions
- range imaging
- multimodal medical images
- multimedia
- scene analysis
- directed acyclic graph
- graph matching
- data fusion
- moving objects
- image segmentation