MAGAN: Margin Adaptation for Generative Adversarial Networks

TL;DR

MAGAN introduces an adaptive hinge loss margin for GAN training, enhancing stability and performance by aligning the margin with the target distribution's energy, leading to improved image generation results.

Contribution

The paper presents a novel margin adaptation method for GANs that dynamically adjusts the hinge loss margin based on the target distribution's energy, with proven convergence guarantees.

Findings

Improved stability and performance in GAN training.

Qualitative and quantitative gains over state-of-the-art methods.

Robustness across diverse datasets.

Abstract

We propose the Margin Adaptation for Generative Adversarial Networks (MAGANs) algorithm, a novel training procedure for GANs to improve stability and performance by using an adaptive hinge loss function. We estimate the appropriate hinge loss margin with the expected energy of the target distribution, and derive principled criteria for when to update the margin. We prove that our method converges to its global optimum under certain assumptions. Evaluated on the task of unsupervised image generation, the proposed training procedure is simple yet robust on a diverse set of data, and achieves qualitative and quantitative improvements compared to the state-of-the-art.