WzyCrush

The test results are different from those in the paper

7

I tested `clean_testset_wav` audio in the Valentini dataset according to the description in your paper. The model used is `vctk_v1`, and the test results are as follows  The test...