Unpaired Image Captioning With semantic-Constrained Self-Learning.
Huixia BenYingwei PanYehao LiTing YaoRichang HongMeng WangTao MeiPublished in: IEEE Trans. Multim. (2022)
Keyphrases
- image data
- image features
- input image
- image classification
- template matching
- image content
- multiscale
- image analysis
- segmentation method
- single image
- semantically meaningful
- natural language
- spatial information
- similarity measure
- image structure
- image collections
- image segmentation
- edge detection
- image retrieval
- semantic categories
- image pixels
- pixel values
- hough transform
- high resolution
- low level
- test images
- denoising
- grey level
- vector field
- image quality
- image matching
- image database
- image regions
- semantic information
- energy function
- medical images
- image representation