Matching Visual Features to Hierarchical Semantic Topics for Image Paragraph Captioning.
Dandan GuoRuiying LuBo ChenZequn ZengMingyuan ZhouPublished in: CoRR (2021)
Keyphrases
- visual features
- image classification
- image collections
- image retrieval
- low level visual features
- semantic gap
- semantic space
- visual appearance
- image similarity
- visual and textual features
- semantic features
- low level features
- visual content
- semantic concepts
- visual similarity
- visual information
- low level
- web images
- visually similar
- global features
- image content
- labeled images
- bridge the semantic gap
- keywords
- bag of features
- image annotation
- key frames
- image search
- visual descriptors
- image matching
- visual data
- image representation
- image features
- sample images
- keypoints
- image data
- visual patterns
- image set
- feature points
- input image
- low level features and high level
- multiscale
- visual concepts
- matching algorithm
- automatic image annotation
- semantic relationships
- feature extraction
- visual properties
- semantically meaningful
- image database
- semantic information
- bag of words
- information retrieval
- textual and visual features
- video retrieval
- sift descriptors
- multi label
- similarity measure
- high level