Protect, detect, mark: This is how DeepMind exposes deceptively real AI-generated fakes
Never be deceived again: SynthID recognizes and marks AI images with an invisible watermark
In a time when artificial intelligence provides us with deceptively real images, the truth is at stake. But Google, through its subsidiary DeepMind, a trump card up its sleeve: SynthID , a groundbreaking tool that will revolutionize the world of digital images.
The Challenge: Real vs. AI-generated images
AI programs like Midjourney are able to generate images that are indistinguishable from reality. This opens the door to potential abuse and the spread of false information. Google is taking action against this danger and has developed SynthID.
SynthID: The AI scanner and watermark generator
How SynthID works is fascinating. It is based on two advanced AI models. The first model is responsible for identifying the images generated. In particular, the metadata is used as an identification feature. The second model deals with watermarking and is able to recognize computer-generated images even if they have been color filtered or resized.
Three levels of trust for maximum security
SynthID rates images in three different trust levels:
- The image is AI generated : an invisible watermark is added here, which is difficult to remove even after editing.
- The image may have been generated and should be used with caution : a warning for those who want to be on the safe side.
- The image was not generated by an AI : Absolute certainty about the authenticity of the image.
SynthID: Still in beta phase
DeepMind emphasizes that SynthID is still in beta and is currently only compatible with its own AI program “Imagen”. However, the future is bright as feedback from users is collected to continually improve the tool. An expansion to other AI programs such as Midjourney or OpenAI's DALL-E 2 is also being considered.
By loading the video, you accept YouTube's privacy policy.
Learn more
Conclusion
The introduction of SynthID is an important step in the fight against the spread of misinformation and the misuse of image files. Google relies on transparency and security to protect the digital world from manipulation.
Stay informed!
Would you like to stay up to date with the latest developments in the digital world? Sign up for the Mimikama newsletter and take advantage of our media education offering to strengthen your digital skills. The truth is in your hands, don't be fooled!
Source:
Deepmind
You might also be interested in:
Lobster sensation or fraud?
Data protection: Meta allows users to control AI model training
The real threat to AI: Not the technology, but who controls it.
Notes:
1) This content reflects the current state of affairs at the time of publication. The reproduction of individual images, screenshots, embeds or video sequences serves to discuss the topic. 2) Individual contributions were created through the use of machine assistance and were carefully checked by the Mimikama editorial team before publication. ( Reason )