Statistical synthesizer with embedded prosodic and spectral modifications to generate highly intelligible speech in noise.
Daniel ErroTudor-Catalin ZorilaYannis StylianouEva NavasInma HernáezPublished in: INTERSPEECH (2013)
Keyphrases
- text to speech
- speech recognition
- text to speech synthesis
- speech synthesis
- prosodic features
- noisy environments
- speech enhancement
- spectral features
- noise level
- linear prediction
- signal to noise ratio
- statistical analysis
- statistical models
- linear predictive coding
- random noise
- speech signal
- vocal tract
- noise model
- statistical information
- embedded systems
- statistical methods
- missing data
- recognition engine
- pattern recognition
- background noise
- automatically generate
- spectral analysis
- spoken language
- noise reduction
- audio visual
- information theoretic
- spectral resolution
- human computer interaction
- speaker adaptation
- language model
- hidden markov models