-
Notifications
You must be signed in to change notification settings - Fork 2.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add DCCRN separator #4097
Add DCCRN separator #4097
Conversation
…on file of dccrn for egs2/dns_ins20
…on file of dccrn for egs2/dns_ins20
… num_spk and use_noise_mask option
… num_spk and use_noise_mask option
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks! Could you add a unit test for DCCRN in test/espnet2/enh? e.g. test/espnet2/enh/test_espnet_model.py and test/espnet2/enh/separator
… explicitly warn the users do not set init method
Codecov Report
@@ Coverage Diff @@
## master #4097 +/- ##
==========================================
- Coverage 80.94% 80.43% -0.51%
==========================================
Files 435 442 +7
Lines 37651 38557 +906
==========================================
+ Hits 30477 31015 +538
- Misses 7174 7542 +368
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
…nh_dccrn.yaml into egs2/dns_ins20/enh1/README.md
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Many thanks!
This is a great job!
I have some minor comments.
Also, you may need to merge the latest PR to pass the CI.
@@ -0,0 +1,436 @@ | |||
import torch |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I missed this function.
This is really cool!
Please add the description about dccrn and this new function in https://github.com/espnet/espnet#se-speech-enhancement-and-separation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I have added the DCCRN and the link of the original paper in the README
- pytorch version: `pytorch 1.9.1` | ||
- Git hash: `6f66283b9eed7b0d5e5643feb18d8f60118a4afc` | ||
- Commit date: `Mon Dec 13 15:30:29 2021 +0800` | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you upload a model to the HF hub and add a link here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, I have uploaded the model to the HF hub, and the link is https://huggingface.co/Johnson-Lsx/Shaoxiong_Lin_dns_ins20_enh_enh_train_enh_dccrn_raw
…_dccrn.yaml, add PESQ and SI_SNR scores in egs2/dns_ins20/enh1/README.md
…ain_enh_dccrn.yaml
Thanks a lot!!! |
Add espnet2/enh/layers/complexnn.py and espnet2/enh/separator/dccrn_separator.py and modified espnet2/tasks/enh.py to implement DCCRN separator, add a configuration file of dccrn for egs2/dns_ins20.