Press "Enter" to skip to content

Adobe Looking To Develop A Tool That Recognizes Edited Images Through Machine Learning

Experts around the world are increasingly concerned about new artificial intelligence tools that make it easier to edit videos and images, especially when you can swap content with social media with shocking speed and accuracy. Some of these tools have been developed by Adobe, but the company is also working on an antidote and examines how machine learning can be used to automatically select edited images.

The latest work from the company, presented this month at the CVPR Computer Vision conference, shows how people-driven digital forensic data can be automated in less time by machines. Research work is not a breakthrough in this area and is not yet available as a commercial product, but it is interesting to see that Adobe focusing on that area for work.

The spokesperson for the company said it was a ‘research project at an early stage’, but the company wants to play a role in the development of technology to check and verify the authenticity of digital media. What this can mean is incomprehensible as Adobe has never produced software designed to detect fake images. But the company emphasizes its work with law enforcement agencies using digital expertise to assist in finding missing children, as evidence of its responsible approach towards its technologies.

The new research demonstrates how machine learning can be deployed to recognize three common forms of image manipulation, splicing, when two parts of the image are combined; cloning, when objects in the image are copied and pasted; and delete when the object has been completely modified through editing.

The software giant is also making the Photoshop accessible on iPad following the app strategy shift. Adobe system is planning to launch the new app at the annual MAX creative conference that will be held in October.

Be First to Comment

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.