Automatic recognition of connected vowels only using speaker-invariant representation of speech dynamics.
Satoshi AsakawaNobuaki MinematsuKeikichi HirosePublished in: INTERSPEECH (2007)
Keyphrases
- automatic recognition
- invariant representation
- speech recognition
- speaker dependent
- vocal tract
- speech sounds
- speech signal
- automatic speech recognition
- speaker identification
- invariant representations
- speech synthesis
- speech recognition systems
- speaker recognition
- affine transformation
- geometric structure
- selective attention
- audio visual
- hidden markov models
- pattern recognition
- speaker diarization
- mel frequency cepstral coefficients
- noisy environments
- speaker verification
- speaker independent
- language model
- speaker adaptation
- formant frequencies
- license plate
- gaussian mixture model
- broadcast news
- speech recognizer
- face recognition