Joint Phoneme Alignment and Text-Informed Speech Separation on Highly Corrupted Speech.
Kilian Schulze-ForsterClement S. J. DoireGaël RichardRoland BadeauPublished in: ICASSP (2020)
Keyphrases
- speech recognition
- speech synthesis
- text to speech
- automatic speech recognition
- text to speech synthesis
- speech signal
- multi lingual
- automatic speech recognition systems
- speaker identification
- speaker dependent
- prosodic features
- english text
- hidden markov models
- phoneme recognition
- spontaneous speech
- recognition engine
- spoken language
- broadcast news
- text input
- text mining
- text recognition
- noisy environments
- lexical features
- visual speech
- speaker recognition
- information retrieval
- language acquisition
- speech corpus
- spoken documents
- audio visual
- speech sounds
- pattern recognition
- keywords