Speech Intelligibility Based Enhancement System Using Modified Deep Neural Network and Adaptive Multi-band Spectral Subtraction.
Tusar Kanti DashSandeep Singh SolankiPublished in: Wirel. Pers. Commun. (2020)
Keyphrases
- spectral subtraction
- speech recognition
- multi band
- speech signal
- hearing aids
- neural network
- hybrid method
- automatic speech recognition
- hidden markov models
- pattern recognition
- speech synthesis
- multispectral
- noisy environments
- frequency band
- language model
- background noise
- vocal tract
- signal to noise ratio
- hyperspectral
- image quality
- edge detection
- speech enhancement