Using Privacy-Transformed Speech in the Automatic Speech Recognition Acoustic Model Training.
Askars SalimbajevsPublished in: Baltic HLT (2020)
Keyphrases
- automatic speech recognition
- speech recognition
- speech signal
- discriminative training
- word error rate
- acoustic models
- broadcast news
- hidden markov models
- conversational speech
- spoken words
- spontaneous speech
- speech recognizers
- speech corpus
- acoustic features
- noisy environments
- word recognition
- speech sounds
- speech recognizer
- speech segments
- speech synthesis
- compound words
- speech retrieval
- language model
- vocal tract
- speaker independent
- phoneme recognition