Enhancing Image Captioning with Transformer-Based Two-Pass Decoding Framework.
Jindian SuYueqi MouYunhao XiePublished in: ICIC (LNAI 1) (2024)
Keyphrases
- image data
- image content
- image analysis
- multiscale
- image features
- template matching
- image classification
- image representation
- similarity measure
- image retrieval
- vector field
- single image
- image segmentation
- bayesian framework
- decoding process
- region of interest
- low level
- image database
- fuzzy logic
- digital images
- registration framework
- image pixels
- pixel values
- image structure
- image collections
- lighting conditions
- segmentation algorithm
- input image
- edge detection
- image matching
- spatial information
- hough transform
- image regions
- segmentation method