Multi resolution discriminative models for subvocalic speech recognition.
Mark RaugasVivek Kumar Rangarajan SridharRohit PrasadPrem NatarajanPublished in: INTERSPEECH (2010)
Keyphrases
- speech recognition
- discriminative models
- multiresolution
- generative model
- language model
- language modeling
- semi supervised learning
- probabilistic model
- discriminative learning
- hidden markov models
- automatic speech recognition
- learning tasks
- conditional random fields
- speech signal
- speech recognizer
- image statistics
- speech recognition systems
- conditional probabilities
- speech synthesis
- n gram
- pattern recognition
- semi supervised
- query expansion
- feature vectors
- complex domains
- learning algorithm
- document retrieval
- em algorithm
- graphical models
- information extraction
- training data
- image processing
- information retrieval
- neural network