Repository navigation
Questions about the VR harmonic-noise separation model #342
Unanswered
wavtechyukky
asked this question in
Q&A
Replies: 2 comments
|
使用了 DDSP 合成,从公开数据集启动,且此类模型不涉及演绎,因此无约束。这个权重没什么使用条款,随便使用。 |
0 replies
|
谢谢!!我觉得这是一个非常值得尝试各种应用的模型。使用时我会注明出处。 |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
感谢您发布 hnsep_240512。想请教以下几个问题:
训练时 harmonic 的目标数据(
instruments/)是如何生成的?例如,是否使用 DDSP 类声码器(如 pc-ddsp / DDSP-SVC 中的 CombSub)分别合成了 harmonic 和 noise?还是使用某种分析方法,从真实录音中分离得到的?训练所使用的数据集,是否一般允许用于研究用途或非商业用途?
请问 hnsep_240512 权重的使用条款是什么?是否允许非营利的研究用途和非商业用途?例如,是否允许使用其输出数据来训练其他模型(如声码器或声学模型)?
任何信息都非常感谢。拜托您了!
Thank you for releasing hnsep_240512. I'd like to ask a few questions:
How were the harmonic targets (
instruments/) generated? For example, were harmonic and noise synthesized separately with a DDSP-style vocoder (e.g. CombSub in pc-ddsp / DDSP-SVC)? Or were real recordings separated using some kind of analysis method?Are the datasets used for training generally licensed for research or non-commercial use?
What are the terms of use for the hnsep_240512 weights? Is non-profit research use or non-commercial use permitted? For example, is it permitted to train other models (such as vocoders or acoustic models) on its outputs?
Any information would be greatly appreciated. Thank you very much!
All reactions