Exemplar-based noise robust automatic speech recognition using modulation spectrogram features.
Deepak BabyTuomas VirtanenJort F. GemmekeTom BarkerHugo Van hammePublished in: SLT (2014)
Keyphrases
- noisy environments
- speech signal
- image noise
- speech recognition
- feature vectors
- automatic speech recognition
- co occurrence
- feature extraction
- salt pepper
- hidden markov models
- speech enhancement
- image features
- noise reduction
- spectral features
- single channel
- graph matching
- signal to noise ratio
- feature set
- classification accuracy
- multiscale