VRSO: Visual-Centric Reconstruction for Static Object Annotation.
Chenyao YuYingfeng CaiJiaxin ZhangHui KongWei SuiCong YangPublished in: CoRR (2024)
Keyphrases
- visual objects
- visual properties
- object reconstruction
- high resolution
- shape reconstruction
- spatial relations
- visual information
- three dimensional
- category specific
- real world objects
- object model
- object tracking
- moving objects
- d objects
- semantic annotation
- contextual cues
- object description
- active learning
- visual scene
- back projection
- automatic image annotation
- image reconstruction
- image annotation
- data objects
- visual appearance
- image representation
- shape recovery
- multiple objects
- keypoints
- image retrieval
- metadata
- web image search