A Spectral Masking Approach to Noise-Robust Speech Recognition Using Deep Neural Networks.
Bo LiKhe Chai SimPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2014)
Keyphrases
- neural network
- noisy environments
- image noise
- pattern recognition
- salt pepper
- speech recognition
- linear predictive coding
- speech enhancement
- genetic algorithm
- greater robustness
- weak edges
- fault diagnosis
- artificial neural networks
- multi layer
- speech signal
- estimation error
- noise level
- spectral analysis
- spectral features
- low frequency
- recurrent neural networks
- noisy data