Towards Accurate Text-Based Image Captioning With Content Diversity Exploration.
Guanghui XuShuaicheng NiuMingkui TanYucheng LuoQing DuQi WuPublished in: CVPR (2021)
Keyphrases
- image data
- image content
- single image
- image retrieval
- input image
- image features
- image classification
- image analysis
- template matching
- image collections
- image pixels
- image representation
- image segmentation
- multimedia
- low level
- multiscale
- pixel values
- high fidelity
- web images
- textual descriptions
- image set
- low level image features
- vector field
- hough transform
- image regions
- test images
- image matching
- image structure
- image search
- spatial information
- segmentation method
- similarity measure
- completely automatic
- image processing