IMLHF / SE_DCUNet

Deep Complex UNet for speech enhancement, init from "https://github.com/chanil1218/DCUnet.pytorch"
13 stars 2 forks source link

se_dataset.py #2

Closed njusq closed 3 years ago

njusq commented 4 years ago

Hi! Thank you for your code. I tried to run your code with my own dataset, and found you changed line 96 in se_dataset.py from input_mat[e, :inp.shape[0]] = inp to input_mat[e] = inp # no padding

Is there anything special in your data so you wrote it like this?

IMLHF commented 3 years ago

i repeat the wav to a fixed length, instead of padding