Injecting Prior Knowledge into Image Caption Generation.
Arushi GoelBasura FernandoThanh-Son NguyenHakan BilenPublished in: ECCV Workshops (2) (2020)
Keyphrases
- prior knowledge
- image data
- image features
- input image
- single image
- multiscale
- image content
- bayesian framework
- image retrieval
- image analysis
- template matching
- image representation
- similarity measure
- prior information
- edge detection
- visual features
- segmentation method
- image classification
- image segmentation
- image structure
- edge map
- feature points
- image regions
- high resolution
- image collections
- video retrieval
- image pixels
- caption text