You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi!
I noticed this description in the deepfilternetv1 paper, but I can't find such an operation in the current code. Does the algorithm still support the clean target of speech with a little noise? Because when I train a small model and perform inference, even if the original noisy is mixed, there will be slight distortion.
Thanks for your work!
The text was updated successfully, but these errors were encountered:
Hi!
I noticed this description in the deepfilternetv1 paper, but I can't find such an operation in the current code. Does the algorithm still support the clean target of speech with a little noise? Because when I train a small model and perform inference, even if the original noisy is mixed, there will be slight distortion.
Thanks for your work!
The text was updated successfully, but these errors were encountered: