Model-based adaptive pre-processing of speech for enhanced intelligibility in noise and reverberation.
Jan RenniesAndreas VolgenandtHenning F. SchepkerSimon DocloPublished in: INTERSPEECH (2015)
Keyphrases
- preprocessing
- speech recognition
- signal to noise ratio
- noisy environments
- data driven
- speech signal
- speech enhancement
- noise level
- noise reduction
- noisy data
- switching median filter
- noise suppression
- post processing
- feature extraction
- speech synthesis
- spatially adaptive
- additive noise
- high noise
- audio visual
- pattern recognition
- noise removal
- noise free
- language acquisition
- random noise
- spoken language
- information retrieval
- noisy images
- missing data
- color images
- feature selection