Recognition rate estimation based on word alignment network and discriminative error type classification.
Atsunori OgawaTakaaki HoriAtsushi NakamuraPublished in: SLT (2012)
Keyphrases
- recognition rate
- classification rate
- word alignment
- structured output
- discriminative learning
- face recognition
- classification accuracy
- feature selection
- feature extraction
- document classification
- feature space
- classification algorithm
- support vector machine svm
- machine learning
- machine translation
- discriminatory power
- error rate
- text classification
- image classification
- training set
- training data
- decision trees
- class labels
- markov random field
- document clustering
- object detection
- feature vectors
- markov networks
- statistical machine translation
- support vector