Investigating NMF speech enhancement for neural network based acoustic models.
Jürgen T. GeigerJort F. GemmekeBjörn W. SchullerGerhard RigollPublished in: INTERSPEECH (2014)
Keyphrases
- speech enhancement
- acoustic models
- speech recognition
- speech signal
- noisy environments
- automatic speech recognition
- noise reduction
- single channel
- hidden markov models
- linear prediction
- neural network
- signal to noise ratio
- speech recognizer
- broadcast news
- speaker independent
- vocal tract
- background noise
- language model
- pattern recognition
- multi channel
- speaker identification
- wiener filter
- smoothing algorithm
- spectral analysis
- discriminative training
- acoustic features
- additive noise
- speech synthesis
- non stationary
- signal processing
- training data