Most deep noise suppression (DNS) models are trained with reference-base...
Wideband codecs such as AMR-WB or EVS are widely used in (mobile) speech...
Perceptual evaluation of speech quality (PESQ) requires a clean speech
r...
Speech enhancement employing deep neural networks (DNNs) for denoising a...
Estimating time-frequency domain masks for single-channel speech enhance...