This is Niche Gamer Tech. In this column, we regularly cover tech and things related to the tech industry. Please leave feedback and let us know if there’s tech or a story you want us to cover!
Adobe and UC Berkeley have developed a new tool that helps determine if someone has touched up a photograph of their face, with Photoshop.
While the tool remains unnamed, it is reportedly twice as effective at identifying Photoshopped images as regular humans can.
This is done via machine learning, where the deep-learning tool is trained on thousands of images gathered from the internet. The new tool spotted ‘shopped photos 99% of the time, while humans were only effective 53% of the time.
The caveat is this tool is only applicable to photos altered with the popular “Face Aware Liquify” tool, which lets you easily and quickly touch up blemishes or unwanted features in a normal portrait shot of a face.
Despite this being just a prototype, Adobe wants to expand this research and develop further tools to help identify and discourage faked or doctored photos – most especially with the danger of “deepfake” photos or videos arising via machine-learning based tools.