BEV2PR: BEV-Enhanced Visual Place Recognition with Structural Cues.
Fudong GeYiwei ZhangShuhan ShenYue WangWeiming HuJin GaoPublished in: CoRR (2024)
Keyphrases
- visual cues
- mid level
- contextual cues
- structural information
- visual features
- visual perception
- low level
- high level
- databases
- visual tasks
- visual patterns
- information retrieval
- structural features
- structural model
- visual representation
- video search
- real world
- three dimensional
- computer vision
- information systems
- visual information
- data sets
- multimedia
- prior knowledge
- case study