Cross database training of audio-visual hidden Markov models for phone recognition.
Shahram KalantariDavid DeanHouman GhaemmaghamiSridha SridharanClinton FookesPublished in: INTERSPEECH (2015)
Keyphrases
- hidden markov models
- audio visual
- multi stream
- acoustic models
- discriminative training
- multi modal
- speaker independent
- baum welch
- gesture recognition
- handwritten text recognition
- speech recognition
- visual information
- object recognition
- multimedia
- visual data
- visual speech recognition
- emotion recognition
- pattern recognition
- minimum classification error
- audio features
- training set
- visual speech
- handwriting recognition
- automatic speech recognition
- activity recognition
- action recognition
- feature extraction