Stabilizing GAN Training with Multiple Random Projections

Abstract

Training generative adversarial networks is unstable in high-dimensions as the true data distribution tends to be concentrated in a small fraction of the ambient space. The discriminator is then quickly able to classify nearly all generated samples as fake, leaving the generator without meaningful gradients and causing it to deteriorate after a point in training. In this work, we propose training a single generator simultaneously against an array of discriminators, each of which looks at a different random low-dimensional projection of the data. Individual discriminators, now provided with restricted views of the input, are unable to reject generated samples perfectly and continue to provide meaningful gradients to the generator throughout training. Meanwhile, the generator learns to produce samples consistent with the full data distribution to satisfy all discriminators simultaneously. We demonstrate the practical utility of this approach experimentally, and show that it is able to produce image samples with higher quality than traditional training with a single discriminator.

References in corpus (6)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
• Improved Techniques for Training GANs
• Energy-based Generative Adversarial Network
• Semi-Supervised Learning with Context-Conditional Generative Adversarial Networks
• Ensembles of Generative Adversarial Networks

Cited by in corpus (20)

• Unsupervised Domain Adaptation through Self-Supervision
• Generalizing to unseen domains via distribution matching
• Spatial Evolutionary Generative Adversarial Networks
• Dropout-GAN: Learning from a Dynamic Ensemble of Discriminators
• Coupled VAE: Improved Accuracy and Robustness of a Variational Autoencoder
• Differentially Private Generative Adversarial Networks with Model Inversion
• Joint Air Quality and Weather Prediction Based on Multi-Adversarial Spatiotemporal Networks
• MCL-GAN: Generative Adversarial Networks with Multiple Specialized Discriminators
• Self-Supervised GANs with Label Augmentation
• CDE-GAN: Cooperative Dual Evolution Based Generative Adversarial Network
• Training Generative Adversarial Networks with Adaptive Composite Gradient
• Kernel-Guided Training of Implicit Generative Models with Stability Guarantees
• 3D Human motion anticipation and classification
• Deconstructing Generative Adversarial Networks
• DepthwiseGANs: Fast Training Generative Adversarial Networks for Realistic Image Synthesis
• Synthesizing 3D Shapes from Silhouette Image Collections using Multi-projection Generative Adversarial Networks
• Coupled Generative Adversarial Network for Continuous Fine-grained Action Segmentation
• A Spectral Enabled GAN for Time Series Data Generation
• Continual Learning of Generative Models with Limited Data: From Wasserstein-1 Barycenter to Adaptive Coalescence
• Learning to navigate image manifolds induced by generative adversarial networks for unsupervised video generation