"Convolutional Deep Neural Network and Full Connectivity for Speech Enh" by Ban M. Alameri, Inas Jawad Kadhim et al.
The speech signal that is received in real-time has background noise and reverberations, which have an impact on the quality of speech. Therefore, it is crucial to reduce or eliminate the noise and increase the intel-ligibility and quality of speech signals. In this study, a proposed method that is the most effective and challenging in a low SNR environment for three types of noise are removed, including washing machine, traffic noise, and electric fan noise, and clean speech is recovered. with ...
Convolutional Network Deep Learning Ully Connected Network Ignal To Noise Ratio Snr Speech Enhancement
Source: uow.edu.au