Research

Loss Functions for Image Restoration with Neural Networks

"Loss Functions for Image Restoration with Neural Networks"
Hang Zhao (MIT), Orazio Gallo (NVIDIA), Iuri Frosio (NVIDIA), Jan Kautz (NVIDIA), in IEEE Transactions on Computational Imaging, March 2017
Research Area: Computational Photography
Imaging
Author(s): Hang Zhao (MIT), Orazio Gallo (NVIDIA), Iuri Frosio (NVIDIA), Jan Kautz (NVIDIA)
Date: March 2017
Download(s):
 
Abstract: Neural networks are becoming central in several areas of computer vision and image processing and different architectures have been proposed to solve specific problems. The impact of the loss layer of neural networks, however, has not received much attention in the context of image processing: the default and virtually only choice is L2. In this paper, we bring attention to alternative choices for image restoration. In particular, we show the importance of perceptually-motivated losses when the resulting image is to be evaluated by a human observer. We compare the performance of several losses, and propose a novel, differentiable error function. We show that the quality of the results improves significantly with better loss functions, even when the network architecture is left unchanged.