mscipio / mscipio.github.io

0 stars 0 forks source link

post/bss-shogun-python/ #11

Open utterances-bot opened 1 year ago

utterances-bot commented 1 year ago

Blind Source Separation (BSS) with the Shogun Machine Learning Toolbox | MICHELE SCIPIONI

How to do Blind Source Separation (BSS) using algorithms available in the Shogun Machine Learning Toolbox.

https://mscipio.github.io/post/bss-shogun-python/

DevGogoi12 commented 1 year ago

Nice work sir, but SHOGUN toolkit functions are not working, like shogun.Features, shogun.Converter and multiple errors. If you have any updated post please share, it will be very helpful.

mscipio commented 1 year ago

Hi! There is a very good chance the toolbox changed over the past 6 years. Have a look here (https://github.com/shogun-toolbox/shogun)and see if you can fill in the gaps of this old post. Maybe I will come up with an updated version of it, but don't wait for it too soon.

DevGogoi12 commented 1 year ago

Thank you for your reply... I will look into it

navinreddy23 commented 3 months ago

Hi Michele,

I am trying to separate speech from multi-source mics. I see quite a lot of mic-bleed. Will this method be applicable for separating the sources or canceling out mic-bleed.

For e.g. I have a setup with 5 mics. I see mic bleed of Speaker 2 in Mic 5. Using audacity I can clear see that the signal is slightly delayed and has less amplitude. If I mix the signals, will I be able to separate the sources?