Noise-robust automatic speech recognition using a discriminative echo state network.
Mark D. SkowronskiJohn G. HarrisPublished in: ISCAS (2007)
Keyphrases
- noisy environments
- echo state networks
- speech recognition
- semi automatic
- noise reduction
- greater robustness
- image noise
- background noise
- speech signal
- geometric distortions
- automatic speech recognition
- estimation error
- recurrent neural networks
- speech enhancement
- fully automatic
- speaker identification
- imaging artifacts
- median filter
- robust estimation
- image sequences
- noise level
- signal to noise ratio
- maximum likelihood
- high dimensional