Google has created a new tool called SynthID to assist in identifying pictures and artwork that were created using artificial intelligence (AI) algorithms as these systems produce increasingly realistic synthetic pictures.
In the initial stages Google allows this technique only to a small number of Google Cloud users to start the process of detecting AI-generated material. With the help of SynthID, Google can still detect AI-generated photos with invisible digital watermarks even after being cropped or given a completely different color.
Google is testing SynthID with a small number of Google Cloud users using Imagen, a new text-to-image model. The digital watermark created by SynthID is integrated into the pixel data, not like the visual watermarks that are frequently employed nowadays.
The tools work with the help of 2 AI models where one AI tool adds the unnoticeable watermarks and the other detects them. During the internal testing, SynthID accurately detects pictures that are AI-generated.
Read Also:
- AI-driven sensors are the key to the success of the Chandrayaan-3 mission
- Perplexity Announces New Updates About AI Search And Code Llama Cha
- Study Tuned With The Impact Of Demographics On AI Training
News Source: https://shorturl.at/absOY
Feature Image Source: https://shorturl.at/hR035
About Author
