Automatic image annotation based on Gaussian mixture model considering cross-modal correlations.
Dongping TianZhongzhi ShiPublished in: J. Vis. Commun. Image Represent. (2017)
Keyphrases
- gaussian mixture model
- automatic image annotation
- visual similarity
- image annotation
- mixture model
- image content
- feature vectors
- em algorithm
- expectation maximization
- semantic concepts
- multi modal
- maximum likelihood
- feature space
- low level features
- visual features
- semantic gap
- multi label
- image retrieval
- visual words
- information retrieval
- co occurrence
- training data
- high level