Spatial Evolutionary Generative Adversarial Networks

Abstract

Generative adversary networks (GANs) suffer from training pathologies such as instability and mode collapse. These pathologies mainly arise from a lack of diversity in their adversarial interactions. Evolutionary generative adversarial networks apply the principles of evolutionary computation to mitigate these problems. We hybridize two of these approaches that promote training diversity. One, E-GAN, at each batch, injects mutation diversity by training the (replicated) generator with three independent objective functions then selecting the resulting best performing generator for the next batch. The other, Lipizzaner, injects population diversity by training a two-dimensional grid of GANs with a distributed evolutionary algorithm that includes neighbor exchanges of additional training adversaries, performance based selection and population-based hyper-parameter tuning. We propose to combine mutation and population approaches to diversity improvement. We contribute a superior evolutionary GANs training method, Mustangs, that eliminates the single loss function used across Lipizzaner's grid. Instead, each training round, a loss function is selected with equal probability, from among the three E-GAN uses. Experimental analyses on standard benchmarks, MNIST and CelebA, demonstrate that Mustangs provides a statistically faster training method resulting in more accurate networks.

References in corpus (15)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• Energy-based Generative Adversarial Network
• Wasserstein GAN
• Towards Principled Methods for Training Generative Adversarial Networks
• Learning What and Where to Draw
• AdaGAN: Boosting Generative Models
• Triangle Generative Adversarial Networks
• Stabilizing GAN Training with Multiple Random Projections
• Ensembles of Generative Adversarial Networks
• Generative Adversarial Parallelization
• Towards Distributed Coevolutionary GANs
• Evolutionary Generative Adversarial Networks
• Lipizzaner: A System That Scales Robust Generative Adversarial Network Training
• Distributional Adversarial Networks

Cited by in corpus (10)

• Random Error Sampling-based Recurrent Neural Network Architecture Optimization
• Exploring the Evolution of GANs through Quality Diversity
• Re-purposing Heterogeneous Generative Ensembles with Evolutionary Computation
• Neural Architecture Search for Generative Adversarial Networks: A Comprehensive Review and Critical Analysis
• Evolving SimGANs to Improve Abnormal Electrocardiogram Classification
• CDE-GAN: Cooperative Dual Evolution Based Generative Adversarial Network
• Multi-population GAN Training: Analyzing Co-Evolutionary Algorithms
• Evolutionary Generative Adversarial Networks with Crossover Based Knowledge Distillation
• Host-Pathongen Co-evolution Inspired Algorithm Enables Robust GAN Training
• Fostering Diversity in Spatial Evolutionary Generative Adversarial Networks