On the Limitations of Visual-Semantic Embedding Networks for Image-to-Text Information Retrieval.
Yan GongGeorgina CosmaHui FangPublished in: J. Imaging (2021)
Keyphrases
- information retrieval
- web images
- semantic space
- input image
- image data
- image content
- semantic content
- single image
- linguistic analysis
- image features
- visual concepts
- text mining
- semantic labels
- high resolution
- image retrieval
- text retrieval
- visual similarity
- multiscale
- computational linguistics
- similarity measure
- semantic information
- image classification
- visual perception
- latent semantic analysis
- low level
- auto annotation
- visual data
- visual appearance
- semantic network
- textual and visual information
- textual descriptions
- visually similar
- web image search
- semantic context
- image collections
- image representation
- information retrieval systems
- natural language
- high level
- image segmentation
- semantic gap
- video search
- pixel values
- image search
- question answering
- visual features
- co occurrence
- digital images
- arabic text