Google is testing watermark to identify AI-generated images: Here’s how it works, what it means for users, and more

Google's AI arm, DeepMind, is testing a digital watermark called SynthID to combat disinformation. The watermark is embedded into AI-generated images, allowing for identification purposes. It has three confidence levels: "Digital watermark detected," "Digital watermark not detected," and "Digital watermark possibly detected." SynthID combines watermarking and identification using metadata and deep learning models. Unlike traditional watermarks, SynthID remains detectable even after modifications. The goal is to reduce misinformation by helping users determine if an image is real or fake.

from Gadgets Now https://ift.tt/4s10Frc
via IFTTT