Comparative experiments to evaluate the use of auditory-based acoustic distinctive features and formant cues for automatic speech recognition using a multi-stream paradigm.
Hesham TolbaSid-Ahmed SelouaniDouglas D. O'ShaughnessyPublished in: INTERSPEECH (2002)
Keyphrases
- multi stream
- distinctive features
- audio visual
- audio visual speech recognition
- speech recognition
- hidden markov models
- sound source
- speech signal
- visual speech
- prosodic features
- speech recognition systems
- visual information
- speech synthesis
- noisy environments
- contextual information
- multimedia
- image data
- acoustic signal
- speech sounds
- vocal tract
- speaker independent
- pattern recognition