Hello,
inspired by the fxDeFeedback in OpenX32 and Chris very good dev blog ( “Neural Networks in Live-Audio-Plugins for De-Feedback“ ) I created a project to prevent Feedback in live environments. I trained a model to detect feedback with it’s main frequency and a plugin with can attack with 8 notch filter to it.
At the moment this is more a draft, but I wan’t to get more training data to it. In future I will create a website where it’s possible to upload recorded feedback and tag it to it’s main frequency.
Feel free to take a look and test it: GitHub - Marvo2011/OpenDeFeedback: AI-powered feedback detection and suppression - ported from OpenMixerProject · GitHub
Maybe feel free to contribute it.