Prescribed Generative Adversarial Networks

Abstract

Generative adversarial networks (GANs) are a powerful approach to unsupervised learning. They have achieved state-of-the-art performance in the image domain. However, GANs are limited in two ways. They often learn distributions with low support---a phenomenon known as mode collapse---and they do not guarantee the existence of a probability density, which makes evaluating generalization using predictive log-likelihood impossible. In this paper, we develop the prescribed GAN (PresGAN) to address these shortcomings. PresGANs add noise to the output of a density network and optimize an entropy-regularized adversarial loss. The added noise renders tractable approximations of the predictive log-likelihood and stabilizes the training procedure. The entropy regularizer encourages PresGANs to capture all the modes of the data distribution. Fitting PresGANs involves computing the intractable gradients of the entropy regularization term; PresGANs sidestep this intractability using unbiased stochastic estimates. We evaluate PresGANs on several datasets and found they mitigate mode collapse and generate samples with high perceptual quality. We further found that PresGANs reduce the gap in performance in terms of predictive log-likelihood between traditional GANs and variational autoencoders (VAEs).

Code for this paper can be found at https://github.com/adjidieng/PresGANs

References in corpus (25)

• Adam: A Method for Stochastic Optimization
• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• Large Scale GAN Training for High Fidelity Natural Image Synthesis
• Trust Region Policy Optimization
• Adversarial Feature Learning
• Adversarially Learned Inference
• Unrolled Generative Adversarial Networks
• Generative Moment Matching Networks
• Variational Approaches for Auto-Encoding Generative Adversarial Networks
• Generalization and Equilibrium in Generative Adversarial Nets (GANs)
• Wasserstein Auto-Encoders
• Learning in Implicit Generative Models
• Demystifying MMD GANs
• Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks
• Learning Generative Models with Sinkhorn Divergences
• Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models
• Maximum Entropy Generators for Energy-Based Models
• Avoiding Latent Variable Collapse With Generative Skip Models
• Metropolis-Hastings Generative Adversarial Networks
• Reweighted Expectation Maximization
• Improving GAN Training via Binarized Representation Entropy (BRE) Regularization
• Learning Implicit Generative Models with the Method of Learned Moments
• Semi-Implicit Generative Model
• Out-of-Sample Testing for GANs

Cited by in corpus (19)

• LOGAN: Latent Optimisation for Generative Adversarial Networks
• Generative models for sampling and phase transition indication in spin systems
• Adversarial score matching and improved sampling for image generation
• Forward and inverse reinforcement learning sharing network weights and hyperparameters
• Probabilistic Time Series Forecasting with Structured Shape and Temporal Diversity
• Top-k Training of GANs: Improving GAN Performance by Throwing Away Bad Samples
• Tail of Distribution GAN (TailGAN): Generative-Adversarial-Network-Based Boundary Formation
• Group Equivariant Generative Adversarial Networks
• VAEBM: A Symbiosis between Variational Autoencoders and Energy-based Models
• No MCMC for me: Amortized sampling for fast and stable training of energy-based models
• GANs with Variational Entropy Regularizers: Applications in Mitigating the Mode-Collapse Issue
• Dual Contradistinctive Generative Autoencoder
• On Out-of-distribution Detection with Energy-based Models
• Alleviation of Gradient Exploding in GANs: Fake Can Be Real
• IID-GAN: an IID Sampling Perspective for Regularizing Mode Collapse
• Imitation with Neural Density Models
• TaylorGAN: Neighbor-Augmented Policy Update for Sample-Efficient Natural Language Generation
• Tractable loss function and color image generation of multinary restricted Boltzmann machine
• Generating Multi-type Temporal Sequences to Mitigate Class-imbalanced Problem