ALGONQUIN - Learning Dynamic Noise Models From Noisy Speech for Robust Speech Recognition.
Brendan J. FreyTrausti T. KristjanssonLi DengAlex AceroPublished in: NIPS (2001)
Keyphrases
- image processing
- speech recognition
- noisy environments
- noisy speech
- pattern recognition
- spectral subtraction
- hidden markov models
- speech signal
- speech enhancement
- speech synthesis
- signal processing
- background noise
- signal to noise ratio
- speaker identification
- image compression
- image restoration
- automatic speech recognition
- language model
- acoustic models
- speech recognition systems
- probabilistic model
- speaker independent
- noise level
- neural network