Weakly-supervised image captioning based on rich contextual information.
Hai-Tao ZhengZhe WangNingning MaJin-Yuan ChenXi XiaoArun Kumar SangaiahPublished in: Multim. Tools Appl. (2018)
Keyphrases
- contextual information
- weakly supervised
- context aware
- image data
- input image
- image features
- image content
- object localization
- multiscale
- image segmentation
- high resolution
- low level
- image classification
- single image
- image retrieval
- superpixels
- test images
- high level
- semantic information
- spatial information
- topic models
- segmentation algorithm
- bounding box
- image representation
- domain specific
- feature vectors