1 paper
Tong Lei, Qinwen Hu, Ziyao Lin +5
The prevailing method for neural speech enhancement predominantly utilizes fully-supervised deep learning with simulated pairs of far-field noisy-reverberant speech and clean speec…