Fully-attentive iterative networks for region-based controllable image and video captioning.
Marcella CorniaLorenzo BaraldiAyellet TalRita CucchiaraPublished in: Comput. Vis. Image Underst. (2023)
Keyphrases
- image segmentation
- image data
- image retrieval
- image analysis
- input image
- multiscale
- image features
- single image
- static images
- image content
- image representation
- video images
- image frames
- edge detection
- image classification
- image collections
- low level
- test images
- video files
- high resolution
- multimedia data
- video data
- segmentation algorithm
- image regions
- spatial domain
- pixel values
- segmentation method
- multimedia
- region based image