Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

#

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Jump to bottom

有关非人声部分 #7

Open
xingyuesama opened this issue Aug 7, 2024 · 0 comments
Open

有关非人声部分 #7

xingyuesama opened this issue Aug 7, 2024 · 0 comments

Comments

@xingyuesama
Copy link

您好，我最近在用你们这个模型进行音高的预测，这个模型的RPA和RCA表现很好，但是OA表现并没有和论文一样那么高，只有65%左右，我用的是MIR-1K训练的模型。 我发现这个模型似乎不能识别非人声部分，所以OA不高。请问有什么方法可以解决这个问题？
需要进行静音训练吗，还是得调参？
另外， 我在论文中看到了你们用CREPE也进行了实验， 我也用了， CREPE也是不能正确识别非人声的，OA贼低，请问你们怎么解决这个问题的？
# for free to join this conversation on GitHub. Already have an account? # to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@xingyuesama