Self-Adaptive Incremental Machine Speech Chain for Lombard TTS with High-Granularity ASR Feedback in Dynamic Noise Condition.
Sashi NovitasariSakriani SaktiSatoshi NakamuraPublished in: ICASSP (2023)
Keyphrases
- text to speech
- automatic speech recognition
- noisy environments
- speech recognition
- low snr
- speech synthesis
- speech signal
- low signal to noise ratio
- word error rate
- high noise
- background noise
- random noise
- noise level
- noise reduction
- prosodic features
- missing data
- wide range
- speech enhancement
- pattern recognition
- spontaneous speech
- noisy data
- broadcast news
- incremental version
- audio visual
- recognition errors
- conversational speech