Distilling Coarse-to-Fine Semantic Matching Knowledge for Weakly Supervised 3D Visual Grounding.
Zehan WangHaifeng HuangYang ZhaoLinjun LiXize ChengYichen ZhuAoxiong YinZhou ZhaoPublished in: CoRR (2023)
Keyphrases
- coarse to fine
- weakly supervised
- semantic matching
- object detection
- multiscale
- multiresolution
- domain knowledge
- object class
- knowledge base
- image registration
- dynamic programming
- visual features
- topic models
- knowledge representation
- natural images
- superpixels
- ontology alignment
- knowledge discovery
- databases
- knowledge sources
- active learning
- web services
- information retrieval