Corrupted input image detection using cycle-consistency-based uncertainty and bias estimators. (IMAGE)

Intelligent Computing

Corrupted input image detection using cycle-consistency-based uncertainty and bias estimators.

Caption

(A) Left: The sharp image (ground truth) and the motion blur kernel used. Right: The noise-corrupted input images and the deblurred outputs. (B) Projection of the data onto the 2D space formed by the 2 most important attributes (cycle-consistency-based uncertainty estimators). (C) Detection accuracy of the new method and 2 baseline methods. (D) Estimated and actual uncertainty and the importance of each attribute for classification.

Credit

Ozcan Lab @UCLA.

Usage Restrictions

Credit must be given to the creator.

License

CC BY

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.