Learning Visual Compound Models from Parallel Image-Text Datasets.
Jan MoringenSven WachsmuthSven J. DickinsonSuzanne StevensonPublished in: DAGM-Symposium (2008)
Keyphrases
- auto annotation
- learning algorithm
- multiscale
- prior knowledge
- low level
- image analysis
- web images
- image data
- single image
- visual effects
- input image
- image features
- learning process
- eye fixations
- supervised learning
- visual appearance
- feature points
- visual patterns
- image content
- learning tasks
- learning models
- random fields
- information retrieval
- image retrieval
- visual attributes
- learned models
- visual perception
- visual cues
- visual data
- segmentation method
- image segmentation
- image quality
- image classification
- edge detection
- object detection
- high resolution