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
Reduce word counts for ESPnet-SE++ Joss paper #4844
Conversation
Codecov Report
@@ Coverage Diff @@
## master #4844 +/- ##
==========================================
+ Coverage 80.63% 80.68% +0.05%
==========================================
Files 536 543 +7
Lines 47523 48233 +710
==========================================
+ Hits 38319 38916 +597
- Misses 9204 9317 +113
Flags with carried forward coverage won't be shown. Click here to find out more.
📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
The overall design looks good to me. |
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 a lot @neillu23! I just have a few minor comments. We can merge the PR after they are resolved.
doc/paper/espnet-se++/paper.md
Outdated
The `forward` function of the class follows the general design in ESPnet2: | ||
|
||
def forward(self, speech_mix, speech_ref, ...) |
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.
Since the description about Trainer
is removed, it might be better to add a hyperlink to the definition of Trainer
below:
which processes speech and only returns losses for `Trainer` to update the model.
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.
updated with this hyperlink:
https://github.com/espnet/espnet/blob/master/espnet2/train/trainer.py#L87-L108
doc/paper/espnet-se++/paper.md
Outdated
@@ -643,56 +227,9 @@ Calling `SeparateSpeech` and `Speech2Text` with unprocessed audios returns the s | |||
|
|||
#### SSE | |||
![](https://i.imgur.com/skZ8uDP.png) |
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.
Could you highlight the python code snippet before creating the screenshot? I think it will make the code more readable.
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.
doc/paper/espnet-se++/paper.md
Outdated
<!-- This API allows the processing of both short audio samples and long audio samples. For long audio samples, you can set the value of arguments `segment_size`, `hop_size` to perform segment-wise SSE on the input speech. --> | ||
<!-- (optionally `normalize_segment_scale` and `show_progressbar`) --> | ||
<!-- Note that the segment-wise processing is disabled by default. --> | ||
|
||
#### Joint-Task | ||
![](https://i.imgur.com/hrj0hJq.png) |
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.
ditto
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.
patch for updating the indent for codes
Thanks, @neillu23! |
No description provided.