dmlguq456 / SepReformer

Official repository of SepReformer for speech separation
Apache License 2.0
97 stars 8 forks source link

WJS- Dataset #1

Open arrrrr3186 opened 3 months ago

arrrrr3186 commented 3 months ago

Hello,

Can you please guide, from where/How can i download the wjs dataset. also I do have the AMI-dataset within me present in my local, is there a way to use it with your repository.

Regards, Anuj

dmlguq456 commented 1 week ago

Thank you very much for your interest and patience. I also sincerely apologize for the delayed response.

Please note that the WSJ0 dataset requires a separate purchase as it is not openly available, and we regret that we are unable to provide further assistance in this regard. As for the AMI dataset, it is mainly used for evaluation purposes as a multi-channel audio dataset, and we anticipate that utilizing it may be somewhat cumbersome.

Instead, we recommend considering the Libri-2Mix dataset, which is based on LibriSpeech and is openly available.