Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities

Abstract

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.

The source codes for both LS-GAN and GLS-GAN are available at \url{https://github.com/maple-research-lab}. LS-GAN is also supported by Microsoft CNTK at \url{https://www.cntk.ai/pythondocs/CNTK_206C_WGAN_LSGAN.html}. The original codes of LS-GAN and GLS-GAN are also available at https://github.com/guojunq/lsgan/ and https://github.com/guojunq/glsgan/

References in corpus (5)

• Conditional Generative Adversarial Nets
• DRAW: A Recurrent Neural Network For Image Generation
• Energy-based Generative Adversarial Network
• C-RNN-GAN: Continuous recurrent neural networks with adversarial training
• Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks

Cited by in corpus (47)

• Spectral Normalization for Generative Adversarial Networks
• How Generative Adversarial Networks and Their Variants Work: An Overview
• Differentially Private Generative Adversarial Network
• Generative Adversarial Networks in Computer Vision: A Survey and Taxonomy
• An empirical study on evaluation metrics of generative adversarial networks
• Towards the Automatic Anime Characters Creation with Generative Adversarial Networks
• Deep Learning in Mobile and Wireless Networking: A Survey
• A survey on GANs for computer vision: Recent research, analysis and taxonomy
• Data-Free Adversarial Distillation
• Deep Bayesian Inversion
• Controlling Physical Attributes in GAN-Accelerated Simulation of Electromagnetic Calorimeters
• Robust Imitation of Diverse Behaviors
• Improving Variational Autoencoder with Deep Feature Consistent and Generative Adversarial Training
• CVAE-GAN: Fine-Grained Image Generation through Asymmetric Training
• Improving the Improved Training of Wasserstein GANs: A Consistency Term and Its Dual Effect
• AET vs. AED: Unsupervised Representation Learning by Auto-Encoding Transformations rather than Data
• Adversarial Training in Affective Computing and Sentiment Analysis: Recent Advances and Perspectives
• Generalization in Generative Adversarial Networks: A Novel Perspective from Privacy Protection
• Comparative Study on Generative Adversarial Networks
• FCC-GAN: A Fully Connected and Convolutional Net Architecture for GANs
• Adversarial Network Bottleneck Features for Noise Robust Speaker Verification
• PointAugment: an Auto-Augmentation Framework for Point Cloud Classification
• Learning Face Age Progression: A Pyramid Architecture of GANs
• On the Effectiveness of Least Squares Generative Adversarial Networks
• Generative Adversarial Networks (GANs): What it can generate and What it cannot?
• UGAN: Untraceable GAN for Multi-Domain Face Translation
• Discriminative Region Proposal Adversarial Networks for High-Quality Image-to-Image Translation
• Pulsar Candidate Identification with Artificial Intelligence Techniques
• Mask-Guided Portrait Editing with Conditional GANs
• Metric Learning-based Generative Adversarial Network
• AVT: Unsupervised Learning of Transformation Equivariant Representations by Autoencoding Variational Transformations
• Data Augmentation for Enhancing EEG-based Emotion Recognition with Deep Generative Models
• Learning Continuous Face Age Progression: A Pyramid of GANs
• Dual Generator Generative Adversarial Networks for Multi-Domain Image-to-Image Translation
• Capturing Variabilities from Computed Tomography Images with Generative Adversarial Networks
• FD-GAN: Face-demorphing generative adversarial network for restoring accomplice's facial image
• Face Aging with Contextual Generative Adversarial Nets
• Depth Structure Preserving Scene Image Generation
• 3D Human motion anticipation and classification
• Robust GANs against Dishonest Adversaries
• Learning Discriminators as Energy Networks in Adversarial Learning
• Alleviation of Gradient Exploding in GANs: Fake Can Be Real
• ReshapeGAN: Object Reshaping by Providing A Single Reference Image
• Generating Relevant Counter-Examples from a Positive Unlabeled Dataset for Image Classification
• Generative Adversarial Networks with Inverse Transformation Unit
• Shared Predictive Cross-Modal Deep Quantization
• Label-Removed Generative Adversarial Networks Incorporating with K-Means