Noise reduction of speech by neural networks and vector quantization.
Kazuo NakataAkihiko SugiuraPublished in: EUROSPEECH (1989)
Keyphrases
- vector quantization
- noise reduction
- neural network
- speech enhancement
- speaker recognition
- noisy environments
- image compression
- signal to noise ratio
- edge detection
- noise level
- neural gas
- vector quantizer
- pattern recognition
- artificial neural networks
- distortion measure
- median filter
- speech recognition
- reduced complexity
- speech signal
- speaker identification
- multilayer perceptron
- codebook design
- automatic speech recognition
- audio visual
- fractal image coding
- hearing aids
- speech synthesis
- neural network model
- multiscale
- text to speech
- vocal tract
- entropy constrained
- noisy images
- back propagation