MSG-GAN: Multi-Scale Gradients for Generative Adversarial Networks

Abstract

While Generative Adversarial Networks (GANs) have seen huge successes in image synthesis tasks, they are notoriously difficult to adapt to different datasets, in part due to instability during training and sensitivity to hyperparameters. One commonly accepted reason for this instability is that gradients passing from the discriminator to the generator become uninformative when there isn't enough overlap in the supports of the real and fake distributions. In this work, we propose the Multi-Scale Gradient Generative Adversarial Network (MSG-GAN), a simple but effective technique for addressing this by allowing the flow of gradients from the discriminator to the generator at multiple scales. This technique provides a stable approach for high resolution image synthesis, and serves as an alternative to the commonly used progressive growing technique. We show that MSG-GAN converges stably on a variety of image datasets of different sizes, resolutions and domains, as well as different types of loss functions and architectures, all with the same set of fixed hyperparameters. When compared to state-of-the-art GANs, our approach matches or exceeds the performance in most of the cases we tried.

arXiv commentCVPR 2020 (Main Conference). Work sponsored by TomTom and Adobe. Code repository: https://github.com/akanimax/msg-stylegan-tf

References in corpus (14)

• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Large Scale GAN Training for High Fidelity Natural Image Synthesis
• Conditional Image Generation with PixelCNN Decoders
• Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
• Improved Training of Wasserstein GANs
• NICE: Non-linear Independent Components Estimation
• Improved Techniques for Training GANs
• Toward Multimodal Image-to-Image Translation
• Which Training Methods for GANs do actually Converge?
• The relativistic discriminator: a key element missing from standard GAN
• Unrolled Generative Adversarial Networks
• LR-GAN: Layered Recursive Generative Adversarial Networks for Image Generation
• Amortised MAP Inference for Image Super-resolution
• Photographic Text-to-Image Synthesis with a Hierarchically-nested Adversarial Network

Cited by in corpus (5)

• From EMBER to FIRE: predicting high resolution baryon fields from dark matter simulations with Deep Learning
• SAG-GAN: Semi-Supervised Attention-Guided GANs for Data Augmentation on Medical Images
• HAD-Net: A Hierarchical Adversarial Knowledge Distillation Network for Improved Enhanced Tumour Segmentation Without Post-Contrast Images
• Faces à la Carte: Text-to-Face Generation via Attribute Disentanglement
• Multi-Attributed and Structured Text-to-Face Synthesis