Statistical acoustic-to-articulatory mapping unified with speaker normalization based on voice conversion.
Hidetsugu UchidaDaisuke SaitoNobuaki MinematsuKeikichi HirosePublished in: INTERSPEECH (2015)
Keyphrases
- speech sounds
- acoustic features
- vocal tract
- prosodic features
- speech recognition
- mel frequency cepstral coefficients
- speech synthesis
- speech signal
- speaker verification
- automatic speech recognition
- text to speech
- data driven
- statistical analysis
- speaker independent
- image acquisition
- acoustic models
- visual features
- speaker identification
- preprocessing
- information theoretic
- hidden markov models
- statistical models
- source localization
- underwater vehicles
- speech recognition systems
- normalization method
- audio features
- music information retrieval