A Sliced Wasserstein Loss for Neural Texture Synthesis

Eric Heitz, Kenneth Vanhoey, Thomas Chambon, Laurent Belcour; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 9412-9420

We address the problem of computing a textural loss based on the statistics extracted from the feature activations of a convolutional neural network optimized for object recognition (e.g. VGG-19). The underlying mathematical problem is the measure of the distance between two distributions in feature space. The Gram-matrix loss is the ubiquitous approximation for this problem but it is subject to several shortcomings. Our goal is to promote the Sliced Wasserstein Distance as a replacement for it. It is theoretically proven, practical, simple to implement, and achieves results that are visually superior for texture synthesis by optimization or training generative neural networks.