Skip to content
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

Tracking static target source among different noise sources at a distance #299

Open
swleungbrian opened this issue Sep 16, 2024 · 5 comments

Comments

@swleungbrian
Copy link

Hi all,

I am trying odas to track static target source among different noise sources as shown below.

problem

Localized noises can be another person speaking next to the target source. Background noises can coherent or diffused as well.

I am using Respeaker v2 and Respeaker HAT for raspberry pi as the mic arrays and got poor separation results.
I tried the same with distance reduced to around 30-50cm and audio separation is much better.
I am not sure if i can tune parameters inside cfg file to fit distance extended to 5+/-2m? if yes, any suggestions on this?

Regards,

Brian

@shenxiaozheng
Copy link

其实这是一个经典的问题,如果speaker距离麦克风阵列较远,那么麦克风阵列接受到的信号中,直达声的成分会较少,而主要成分是反射声和混响。反射声和混响对DOA估计是起反作用的。我认为这个库是无法解决这个问题的。在远场中使用这个库会直接歇菜。

@swleungbrian
Copy link
Author

其实这是一个经典的问题,如果speaker距离麦克风阵列较远,那么麦克风阵列接受到的信号中,直达声的成分会较少,而主要成分是反射声和混响。反射声和混响对DOA估计是起反作用的。我认为这个库是无法解决这个问题的。在远场中使用这个库会直接歇菜。

我有試過把 sst 設為 static, 配以已知的DOA, 但是效果很奇怪. 似乎比不自行輸入DOA效果還差很多.

@shenxiaozheng
Copy link

我也试过,准确率太低。

@swleungbrian
Copy link
Author

我也试过,准确率太低。

那就我的應用場景, 值得投放時間去study odas 嗎? 還是自行重寫? 但我覺得如期我由零開始去試 LCMV/MVDR/GSC 那些方法. 那不如看看odas 能否改改去切合自己的應用場境. 不過odas 的代碼也要花時間去理解. 文件夾都很工整, 但就少有comments. 增加了不少理解上的難度.

@shenxiaozheng
Copy link

这些方法,我全部试过,在大混响场景准确率都不高吧。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants