Speech Enhancement for Speaker Recognition Using Deep Recurrent Neural Networks.
Maxim TkachenkoAlexander YamshininNikolay LyubimovMikhail KotovMarina NastasenkoPublished in: SPECOM (2017)
Keyphrases
- recurrent neural networks
- speech enhancement
- noisy environments
- vocal tract
- speaker verification
- speech signal
- speech recognition
- speaker identification
- noise reduction
- automatic speech recognition
- signal to noise ratio
- speaker recognition
- single channel
- noisy speech
- neural network
- feed forward
- linear prediction
- speech synthesis
- wiener filter
- recurrent networks
- background noise
- nonlinear dynamic systems
- echo state networks
- reservoir computing
- audio visual
- image acquisition
- artificial neural networks
- additive noise
- acoustic features
- non stationary
- smoothing algorithm
- pattern recognition
- edge preserving
- image restoration
- mel frequency cepstral coefficients
- principal component analysis
- machine learning