Controllable Contextualized Image Captioning: Directing the Visual Narrative through User-Defined Highlights.
Shunqi MaoChaoyi ZhangHang SuHwanjun SongIgor ShalyminovWeidong CaiPublished in: CoRR (2024)
Keyphrases
- user defined
- input image
- image analysis
- image data
- image features
- image content
- low level
- multiscale
- visual perception
- image representation
- image retrieval
- single image
- image regions
- image classification
- data types
- visually similar
- image collections
- high resolution
- visual appearance
- visual cues
- test images
- visual attributes
- spatial information
- segmentation method
- image segmentation
- visual information
- segmentation algorithm
- edge detection
- pixel values
- nearest neighbor
- data analysis
- visual concepts
- data structure
- human observers
- databases