Distorted Representation Space Characterization Through Backpropagated Gradients
This work addresses the challenge of handling distorted test image distributions for researchers in computer vision and deep learning, though it is incremental as it builds on existing gradient analysis methods.
The paper tackled the problem of characterizing representation spaces in deep learning by using weight gradients from backpropagation, showing that gradient-based features outperform activation features in perceptual image quality assessment and out-of-distribution classification, achieving top performance on TID 2013 and MULTI-LIVE databases in terms of accuracy, consistency, linearity, and monotonic behavior.
In this paper, we utilize weight gradients from backpropagation to characterize the representation space learned by deep learning algorithms. We demonstrate the utility of such gradients in applications including perceptual image quality assessment and out-of-distribution classification. The applications are chosen to validate the effectiveness of gradients as features when the test image distribution is distorted from the train image distribution. In both applications, the proposed gradient based features outperform activation features. In image quality assessment, the proposed approach is compared with other state of the art approaches and is generally the top performing method on TID 2013 and MULTI-LIVE databases in terms of accuracy, consistency, linearity, and monotonic behavior. Finally, we analyze the effect of regularization on gradients using CURE-TSR dataset for out-of-distribution classification.