Speech Reductions Cause a De-Weighting of Secondary Acoustic Cues.
Léo VarnetFanny MeunierMichel HoenPublished in: INTERSPEECH (2016)
Keyphrases
- prosodic features
- speech synthesis
- text to speech
- speaker verification
- speech sounds
- acoustic features
- speech recognition systems
- speech recognition
- speech recognizers
- emotional speech
- spontaneous speech
- speech signal
- automatic speech recognition
- acoustic signal
- speaker recognition
- speaker independent
- vocal tract
- vowel phonemes
- formant frequencies
- speech recognizer
- noisy environments
- weighting scheme
- acoustic models
- audio visual
- visual features
- noisy speech
- emotion recognition
- efficient computation
- visual cues
- multi modal
- endpoint detection
- spoken dialogue systems
- low level
- hidden markov models
- feature extraction