Home Learn Google DeepMind has launched a watermarking tool for AI-generated images

Google DeepMind has launched a watermarking tool for AI-generated images

0
Google DeepMind has launched a watermarking tool for AI-generated images

Google DeepMind has launched a brand new watermarking tool that labels whether images have been generated with AI.

The tool, called SynthID, will initially be available only to users of Google’s AI image generator Imagen, which is hosted on Google Cloud’s machine learning platform Vertex. Users will find a way to generate images using Imagen after which select whether so as to add a watermark or not. The hope is that it could help people tell when AI-generated content is being passed off as real, or help protect copyright. 

Up to now 12 months, the massive popularity of generative AI models has also brought with it the proliferation of AI-generated deepfakes, nonconsensual porn, and copyright infringements. Watermarking—a way where you hide a signal in a bit of text or a picture to discover it as AI-generated—has turn into one of the vital popular ideas proposed to curb such harms. 

In July, the White House announced it had secured voluntary commitments from leading AI firms reminiscent of OpenAI, Google, and Meta to develop watermarking tools in an effort to combat misinformation and misuse of AI-generated content. 

At Google’s annual conference I/O in May, CEO Sundar Pichai said the corporate is constructing its models to incorporate watermarking and other techniques from the beginning. Google DeepMind is now the primary Big Tech company to publicly launch such a tool.

Traditionally images have been watermarked by adding a visual overlay onto them, or adding information into their metadata. But this method is “brittle” and the watermark might be lost when images are cropped, resized, or edited, says Pushmeet Kohli, vp of research at Google DeepMind.

SynthID is created using two neural networks. One takes the unique image and produces one other image that appears almost equivalent to it, but with some pixels subtly modified. This creates an embedded pattern that’s invisible to the human eye. The second neural network can spot the pattern and can tell users whether it detects a watermark, suspects the image has a watermark, or finds that it doesn’t have a watermark. Kohli said SynthID is designed in a way meaning the watermark can still be detected even when the image is screenshotted or edited—for instance, by rotating or resizing it. 

Google DeepMind just isn’t the just one working on these forms of watermarking methods,  says Ben Zhao, a professor on the University of Chicago, who has worked on systems to stop artists’ images from being scraped by AI systems. Similar techniques exist already and are utilized in the open-source AI image generator Stable Diffusion. Meta has also conducted research on watermarks, even though it has yet to launch any public watermarking tools. 

Kohli claims Google DeepMind’s watermark is more immune to tampering than previous attempts to create watermarks for images, although still not perfectly immune.  

But Zhao is skeptical. “There are few or no watermarks which have proven robust over time,” he says. Early work on watermarks for text has found that they’re easily broken, normally inside just a few months. 

Bad actors have a vested interest in disrupting watermarks, he adds—for instance, to say that deepfaked content is real photographic evidence of a nonexistent crime or event. 

 “An attacker looking for to advertise deepfake imagery as real, or discredit an actual photo as fake, may have quite a bit to realize, and is not going to stop at cropping, or lossy compression or changing colours,” Zhao says. 

Nevertheless, Google DeepMind’s launch is a superb first step and could lead on to higher information-sharing in the sector about which techniques work and which don’t, says Claire Leibowicz, the pinnacle of the AI and Media Integrity Program on the Partnership on AI. 

“The proven fact that this is basically complicated shouldn’t paralyze us into doing nothing,” she says. 

Kohli told MIT Technology Review the watermarking tool is  “experimental” and said the corporate desires to see how people use it and find out about its strengths and weaknesses before rolling it out more widely. He refused to say whether Google DeepMind might make the tool more widely available for images apart from ones generated by Imagen. He also didn’t say whether Google will add the watermark to its AI image generation systems.

This limits its usefulness, says Sasha Luccioni, an AI researcher at startup Hugging Face. Google’s decision to maintain the tool proprietary means only Google will find a way to each embed and detect these watermarks, she adds. 

“In case you add a watermarking component to image generation systems across the board, there shall be less risk of harms like deepfake pornography,” Luccioni says. 

LEAVE A REPLY

Please enter your comment!
Please enter your name here