Joint Discriminative Front End and Back End Training for Improved Speech Recognition Accuracy.
Jasha DroppoAlex AceroPublished in: ICASSP (1) (2006)
Keyphrases
- back end
- recognition accuracy
- recognition rate
- discriminating power
- user friendly
- hearing impaired
- face recognition
- speech recognition
- building blocks
- discriminatively trained
- data management
- data types
- discriminative training
- version control
- improve the recognition rate
- training set
- semi supervised
- speech signal
- automatic speech recognition
- detailed design
- feature extraction
- computer vision
- data streams
- elastic graph matching
- data sets