[ad_1]
San Francisco – Facebook scientists said they developed artificial intelligence software not only to identify “deepfake” images, but also to understand where they came from.
Deepfakes are photos, videos or audio clips that have been altered using artificial intelligence to appear authentic, which experts say can be misleading or completely fake.
Facebook researchers Tal Hassner and Xi Yin said their team worked with Michigan State University to create software that decodes deepfake images to understand how they were made and where they originated.
“Our method will facilitate the detection and tracking of deepfakes in real-world environments, where the deepfake image itself is often the only one information detectors have to work with,” the scientists said in a blog post.
“This work will provide researchers and practitioners with tools to better investigate coordinated disinformation incidents using deepfakes, as well as open new directions for future research,” they added.
Facebook’s new software deepfakes across a network to look for imperfections left during the manufacturing process, which scientists say alter an image’s “fingerprint”.
‘In digital photography, fingerprints are used to identify the digital camera used to produce an image,’ said the scientists.
“Similar to device fingerprints, image fingerprints are unique patterns left on images … which can equally be used to identify the generative pattern from which the image comes.”
“Our research pushes the boundaries of understanding in deepfakes detection,” they said.
Microsoft late last year unveiled software that can help locate deepfake photos or videos, adding to an arsenal of programs designed to combat hard-to-detect images before the US presidential election.
The company’s Video Authenticator software analyzes an image or each frame of a video, looking for evidence of manipulation that may be invisible to the naked eye.
[ad_2]
picture credit