DOCTR: Disentangled Object-Centric Transformer for Point Scene Understanding.
Xiaoxuan YuHao WangWeiming LiQiang WangSoonYong ChoYounghun SungPublished in: AAAI (2024)
Keyphrases
- scene understanding
- object detection
- object recognition
- vision system
- object hypotheses
- motion cues
- scene recognition
- d scene
- robot navigation
- d objects
- video surveillance
- scene categorization
- object segmentation
- scene labeling
- moving objects
- indoor scenes
- scene interpretation
- bag of features
- focus of attention
- image processing
- image parsing
- object tracking
- multi class
- object detectors
- state space
- video sequences
- computer vision
- geometric reasoning
- real time