Researchers at the University of California, Berkeley have developed a new model that enables users to transform images in a "cycle consistent" way, which means any changes to the original image are expected to remain fully reversible.
For example, the algorithm could be used to turn a zebra into a horse, or vice versa.
The researchers say the new model is especially powerful and versatile because it combines the functionalities of several artificial intelligence methods at once. They note this means the algorithm can be easily modified to perform a range of different tasks and transformations.
For example, the researchers say the system can convert paintings into photos as easily as it can transform photos into paintings. However, they also note the model "does not work well when a test image looks unusual compared to training images."
From The Next Web
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA