Thomas Deacon
@adjuvant.bsky.social
Partially meat based system.
For visual AI there are a few data poisoning techniques (Glaze, Nightshade), but we are not there yet for audio. I have seen a couple of things for speech specifically (AntiFake - Yu, Zhai, Zhang 2023), but these might just be research prototypes and not end user tools.
January 15, 2025 at 1:56 PM
For visual AI there are a few data poisoning techniques (Glaze, Nightshade), but we are not there yet for audio. I have seen a couple of things for speech specifically (AntiFake - Yu, Zhai, Zhang 2023), but these might just be research prototypes and not end user tools.
Teenage me is getting all the feels
December 11, 2024 at 5:03 PM
Teenage me is getting all the feels