Ensemble approach based on conditional random field for multi-label image and video annotation.
Xin-Shun XuYuan JiangLiang PengXiangyang XueZhi-Hua ZhouPublished in: ACM Multimedia (2011)
Keyphrases
- multi label
- image classification
- conditional random fields
- segmentation method
- image annotation
- image features
- visual concepts
- multiscale
- image segmentation
- image content
- input image
- text categorization
- image representation
- low level
- graph cuts
- image collections
- multi label classification
- image retrieval
- text classification
- support vector
- test images
- energy function
- graphical models
- visual features
- image regions
- higher order
- bag of words
- probabilistic model
- pairwise
- markov random field
- learning tasks
- image set
- web images
- semantic gap
- computer vision
- machine learning