Abstract:<p>Generative adversarial networks (GANs) have been enjoying considerable success in recent years. However, mode collapse remains a major unsolved problem in training GANs and is one of the main obstacles hindering progress. In this paper, we present spectral regularization for GANs (SR-GANs), a new and robust method for combating the mode collapse problem in GANs. We first perform theoretical analysis to show that the spectral distributions of the weight matrix in the discriminator affect how the equality of Lipschitz constraint can be fulfilled, thus will have an impact on the performance of the discriminator. Prompted by these analysis, we set out to monitor the spectral distributions in the discriminators of spectral normalized GANs (SN-GANs), and discover a phenomenon which we refer to as spectral collapse, where a large number of singular values of the weight matrices drop dramatically when mode collapse occurs. We show that there are strong evidences linking mode collapse to spectral collapse; and based on this link, we set out to tackle spectral collapse as a surrogate of mode collapse. We have developed a spectral regularization method where we introduce two schemes, one static and one dynamic, to compensate the spectral distributions of the weight matrices to prevent them from collapsing, which in turn successfully prevents mode collapse in GANs. Through gradient analysis, we provide theoretical explanations for why SR-GANs are more stable and can provide better performances than SN-GANs. We also present extensive experimental results and analysis to show that SR-GANs not only always outperform SN-GANs but also always succeed in combating mode collapse where SN-GANs fail. The code is available at <a href="https://github.com/max-liu-112/SRGANs-Spectral-Regularization-GANs-"><span class="monospace">https://github.com/max-liu-112/SRGANs-Spectral-Regularization-GANs-</span></a></p>

Mitigating Label Noise in GANs Via Enhanced Spectral Normalization

Spectral Normalization for Generative Adversarial Networks

A Multi-Scale Generative Adversarial Network for Real-World Image Denoising

Why Spectral Normalization Stabilizes GANs: Analysis and Improvements

Incremental Focal Loss GANs.

UGC: Unified GAN Compression for Efficient Image-to-Image Translation

CHAIN: Enhancing Generalization in Data-Efficient GANs via lipsCHitz continuity constrAIned Normalization

Adaptive Spectral Normalization for Generative Models

On Computation and Generalization of GANs with Spectrum Control

Improving GAN Training via Feature Space Shrinkage

Improved Performance of GANs via Integrating Gradient Penalty with Spectral Normalization.

Image enhancement with bi-directional normalization and color attention-guided generative adversarial networks

Generative Adversarial Networks with Denoising Penalty and Sample Augmentation

Learning-aware feature denoising discriminator

A Systematic Survey of Regularization and Normalization in GANs

Towards the Gradient Vanishing, Divergence Mismatching and Mode Collapse of Generative Adversarial Nets

Consistency Regularization for Generative Adversarial Networks

Generative Adversarial Networks with Learnable Auxiliary Module for Image Synthesis

Spectral regularization for combating mode collapse in GANs

Partial Label Supervision for Agnostic Generative Noisy Label Learning

Turning Waste into Wealth: Leveraging Low-Quality Samples for Enhancing Continuous Conditional Generative Adversarial Networks