Improving Image Captioning through Visual and Semantic Mutual Promotion.
Jing ZhangYingshuai XieXiaoqiang LiuPublished in: ACM Multimedia (2023)
Keyphrases
- low level
- single image
- image data
- input image
- image collections
- multiscale
- visual perception
- image features
- semantic space
- visual concepts
- image classification
- image content
- segmentation method
- visually similar
- image analysis
- high level
- visual similarity
- image retrieval
- visual cues
- image matching
- test images
- image segmentation
- high level semantics
- human visual
- computer vision
- visual data
- image representation
- low level features
- visual information
- keypoints
- web images
- high resolution
- visual appearance
- human observers
- visual features
- feature points
- natural language
- scene categorization
- visual attributes
- semantic categories
- edge detection
- low level visual features
- visual patterns
- segmentation algorithm
- semantic gap
- semantic concepts
- image pixels