Speech Enhancement Algorithm Based on a Convolutional Neural Network Reconstruction of the Temporal Envelope of Speech in Noisy Environments.
Rahim SoleymanpourMohammad SoleymanpourAnthony J. BrammerMichael T. JohnsonInsoo KimPublished in: IEEE Access (2023)
Keyphrases
- speech enhancement
- noisy environments
- noise reduction
- speech signal
- signal to noise ratio
- single channel
- linear prediction
- speech recognition
- vocal tract
- noisy speech
- smoothing algorithm
- speaker verification
- additive noise
- background noise
- wiener filter
- speaker identification
- detection algorithm
- convolutional neural network
- spectral subtraction
- pattern recognition
- voice activity detection
- automatic speech recognition
- prediction error
- maximum likelihood
- super resolution
- motion estimation