Unsupervised Speech Enhancement Based on Multichannel NMF-Informed Beamforming for Noise-Robust Automatic Speech Recognition.
Kazuki ShimadaYoshiaki BandoMasato MimuraKatsutoshi ItoyamaKazuyoshi YoshiiTatsuya KawaharaPublished in: CoRR (2019)
Keyphrases
- speech enhancement
- noisy environments
- automatic speech recognition
- speech signal
- speech recognition
- noise reduction
- linear prediction
- spectral subtraction
- single channel
- background noise
- speaker identification
- additive noise
- negative matrix factorization
- conversational speech
- hidden markov models
- vocal tract
- broadcast news
- signal to noise ratio
- multiscale
- acoustic features
- image segmentation
- speech synthesis
- face recognition