Enhancement of noisy speech for noise robust front-end and speech reconstruction at back-end of DSR system.
Hyoung-Gook KimMarkus SchwabNicolas MoreauThomas SikoraPublished in: INTERSPEECH (2003)
Keyphrases
- back end
- noisy speech
- noisy environments
- speech enhancement
- speech recognition
- hidden markov models
- background noise
- speech signal
- signal to noise ratio
- noise reduction
- automatic speech recognition systems
- spectral subtraction
- user friendly
- data management
- data types
- building blocks
- speaker verification
- single channel
- automatic speech recognition
- speech synthesis
- image processing
- low contrast
- linear prediction
- gaussian noise
- database
- wavelet transform
- speaker identification
- language model
- image compression
- image enhancement