Zero-TextCap: Zero-shot Framework for Text-based Image Captioning.
Dongsheng XuWenye ZhaoYi CaiQingbao HuangPublished in: ACM Multimedia (2023)
Keyphrases
- image data
- template matching
- multiscale
- image features
- image classification
- bayesian framework
- image analysis
- vector field
- single image
- image segmentation
- image collections
- image content
- image regions
- image representation
- edge detection
- input image
- low level
- pixel values
- image structure
- image matching
- test images
- object recognition
- image pixels
- pixel level
- grey level
- keypoints
- segmentation method
- feature points
- multi view
- probabilistic model