A neural image captioning model with caption-to-images semantic constructor.
Jinsong SuJialong TangZiyao LuXianpei HanHaiying ZhangPublished in: Neurocomputing (2019)
Keyphrases
- input image
- image data
- image features
- test images
- image classification
- image analysis
- high level
- image dependent
- image set
- geometric information
- image retrieval
- original images
- edge detection
- segmentation algorithm
- image regions
- image pixels
- segmentation method
- image collections
- image description
- high level semantics
- similarity measure
- image statistics
- reconstruction method
- lighting conditions
- low level
- image database
- synthesized images
- restoration method
- multiscale
- sample images
- image similarity
- keypoints
- web images
- bounding box
- visual concepts
- spatial information
- low level features
- reconstruction process
- d objects
- pixel level
- object recognition
- region of interest
- image structure
- image processing