[P] Adversarial Attacks on YouTube Content ID
Does anyone know if it’s possible to do adversarial attacks on YouTube Content ID to make it not detect or detect something else. The most information about the system I could find was 10 years old. See [here]( https://stackoverflow.com/questions/649116/youtube-content-identification-technology). The system is probably very different now.
Any information to help me get started would be awesome.