Combining multiple spatial hidden Markov models in image semantic classification and annotation.
Lihua WangHorace Ho-Shing IpPublished in: ICPR (2008)
Keyphrases
- hidden markov models
- combining multiple
- image classification
- sequence classification
- image retrieval
- spatial information
- image features
- markov models
- multiple classifiers
- image segmentation
- discriminative training
- speech recognition
- discriminative classifiers
- fusion methods
- pattern recognition
- image representation
- hidden state
- image content
- text classification
- conditional random fields
- viterbi algorithm
- spatio temporal
- handwritten words
- feature extraction
- segmentation method
- decision trees
- fusion method
- feature selection
- bag of words
- multiscale
- machine learning
- spatial and temporal
- feature set
- support vector machine
- feature vectors
- training set
- reinforcement learning
- hidden states
- stochastic context free grammars
- baum welch