CSTNET: Enhancing Global-To-Local Interactions for Image Captioning.
Xin YangYing WangHaiShun ChenJie LiPublished in: ICIP (2022)
Keyphrases
- image data
- multiscale
- image analysis
- input image
- image features
- image content
- single image
- image classification
- image pixels
- image collections
- template matching
- image regions
- test images
- low level
- high resolution
- image retrieval
- video sequences
- edge detection
- image representation
- pixel values
- region of interest
- contrast enhancement
- vector field
- image matching
- segmentation method
- pixel level
- global information
- grey level
- computer vision
- global features
- aerial images
- hough transform
- energy function
- similarity measure
- image segmentation