Stack-VS: Stacked Visual-Semantic Attention for Image Caption Generation.
Ling ChengWei WeiXianling MaoYong LiuChunyan MiaoPublished in: IEEE Access (2020)
Keyphrases
- image data
- visual features
- multiscale
- image classification
- single image
- input image
- image features
- image segmentation
- low level
- visual appearance
- image analysis
- visual perception
- high resolution
- visual concepts
- image retrieval
- test images
- bounding box
- image content
- image representation
- auto annotation
- semantic categories
- semantic space
- feature points
- image collections
- semantic labels
- visual cues
- web images
- edge detection
- high level
- low level visual features
- semantic content
- visual data
- segmentation method
- semantic information
- visual patterns
- video retrieval
- visual information
- visually similar
- scene categorization
- high level semantics
- natural language
- similarity measure