Parsing fashion image into mid-level semantic parts based on chain-conditional random fields.
Fan WangQiyang ZhaoBaolin YinXu TaoPublished in: IET Image Process. (2016)
Keyphrases
- natural language processing
- mid level
- natural language
- low level
- visual concepts
- low level features
- high level
- visual cues
- input image
- image data
- single image
- image content
- levels of abstraction
- image classification
- perceptual grouping
- action recognition
- visual information
- semantic information
- image features
- region of interest
- multiscale
- superpixels
- event detection
- image representation
- image patches
- video sequences
- image collections
- visual data
- test images
- segmentation method
- semantic gap
- visual features
- image segmentation