Everyone is increasingly concerned about the fake videos and pictures, but Adobe has come up with the solution where the AI tool can automatically spot photoshopped images. The sharing of new research in collaboration with scientists from UC Berkeley that uses machine learning to automatically detect when images of faces have been manipulated.
Presently the company is committing more resources to this problem.
The Algorithm Spotted 99 Percent Of Edited Faces In Adobe's Tests
Img Src:commsmea
To create the software, engineers trained a neural network on a database of paired faces, containing images both before and after they’d been edited using Liquify.
The resulting algorithm is impressively effective. During the testing phase when asked to spot a sample of edited faces, human volunteers got the right answer 53 per cent of the time, while the algorithm was correct 99 per cent of the time. The tool is even able to suggest how to restore a photo to its original, unedited appearance, though these results are often mixed.
We live in a world where it's harder to trust the digital Information we consume
Img Src:imgflip.com
“The planning of magic universal ‘undo’ button reverts image edits but is still far from reality,” Adobe researcher Richard Zhang, who helped conduct the work, said in a company blog post.
The researchers said the work was the first of its kind designed to spot these sort of facial edits, and constitutes an “important step” toward creating tools that can identify the most complex changes including “body manipulations and photometric edits such as skin smoothing.”
While the research is promising, tools like these are no silver bullet for stopping the harmful effects of manipulated media. As we’ve seen with the spread of fake news, even if the content is obviously false or can be quickly debunked, it will still be shared and embraced on social media. Knowing something is fake is only half the battle, but at least it’s a start.
