Abstract: Deep learning based speech enhancement in the short-time Fourier transform (STFT) domain typically uses a large window length such as 32 ms. A larger window can lead to higher frequency ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results