More Grounded Image Captioning by Distilling Image-Text Matching Model.
Yuanen ZhouMeng WangDaqing LiuZhenzhen HuHanwang ZhangPublished in: CoRR (2020)
Keyphrases
- input image
- image data
- image features
- image matching
- matching scheme
- low level
- image analysis
- image content
- keypoints
- bayesian framework
- multiscale
- template matching
- image retrieval
- statistical model
- single image
- pixel values
- image collections
- similarity measure
- segmentation method
- energy function
- random fields
- high level
- matching process
- prior model
- image similarity
- test images
- feature points
- image classification
- region of interest
- probability density function
- image set
- image regions
- image representation
- bounding box
- energy functional
- edge detection
- probabilistic model
- high resolution