Describing like humans: on diversity in image captioning.
Qingzhong WangAntoni B. ChanPublished in: CoRR (2019)
Keyphrases
- input image
- image features
- image data
- single image
- image content
- multiscale
- template matching
- segmentation method
- image retrieval
- image collections
- image segmentation
- image structure
- image classification
- image representation
- image matching
- spatial information
- lighting conditions
- human observers
- motion estimation
- image analysis
- image regions
- keypoints
- post processing
- region of interest
- low level
- high resolution
- object recognition
- human behavior
- image pixels
- pixel level
- image noise
- human vision
- similarity measure