Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities

• Published in: International journal of computer vision Vol. 128; no. 5; pp. 1118 - 1140
• Main Author: Qi, Guo-Jun
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.05.2020; Springer; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; Computer Imaging; Computer Science; Generative adversarial networks; Image classification; Image Processing and Computer Vision; Image reconstruction; Pattern Recognition; Pattern Recognition and Graphics; Regularization; Semi-supervised learning; Vision
• ISSN: 0920-5691; 1573-1405
• DOI: 10.1007/s11263-019-01265-2

In this paper, we present the Lipschitz regularization theory and algorithms for a novel Loss-Sensitive Generative Adversarial Network (LS-GAN). Specifically, it trains a loss function to distinguish between real and fake samples by designated margins, while learning a generator alternately to produce realistic samples by minimizing their losses. The LS-GAN further regularizes its loss function with a Lipschitz regularity condition on the density of real data, yielding a regularized model that can better generalize to produce new data from a reasonable number of training examples than the classic GAN. We will further present a Generalized LS-GAN (GLS-GAN) and show it contains a large family of regularized GAN models, including both LS-GAN and Wasserstein GAN, as its special cases. Compared with the other GAN models, we will conduct experiments to show both LS-GAN and GLS-GAN exhibit competitive ability in generating new images in terms of the Minimum Reconstruction Error (MRE) assessed on a separate test set. We further extend the LS-GAN to a conditional form for supervised and semi-supervised learning problems, and demonstrate its outstanding performance on image classification tasks.