How to fight the reality of deepfakes
- 5 years ago
AMERICA — Advances in artificial intelligence software has led to the proliferation of fake digital images and videos referred to as "deepfakes."
According to Neimen Lab, deepfakes use several techniques including faceswaps and grafting a lip-syncing mouth or expression onto someone's face.
According to MIT Technology Review, the technology behind deepfakes are called generative adversarial networks, or GANs.
Media outlets like the Wall Street Journal are now training their journalists to detect deepfakes by finding older versions of footage using reverse image search engines like Google Images and Tineye.
Reporters can also use video editing programs to look at clips frame-by-frame to look for distortions in the video.
According to MIT Technology Review, experts say this is just the beginning for GANs. In the future algorithms will become more refined and harder to identify as fakes.
According to Neimen Lab, deepfakes use several techniques including faceswaps and grafting a lip-syncing mouth or expression onto someone's face.
According to MIT Technology Review, the technology behind deepfakes are called generative adversarial networks, or GANs.
Media outlets like the Wall Street Journal are now training their journalists to detect deepfakes by finding older versions of footage using reverse image search engines like Google Images and Tineye.
Reporters can also use video editing programs to look at clips frame-by-frame to look for distortions in the video.
According to MIT Technology Review, experts say this is just the beginning for GANs. In the future algorithms will become more refined and harder to identify as fakes.