Differentiable Augmentation for Data-Efficient GAN Training

• URL: http://arxiv.org/abs/2006.10738v4
• Date: Mon, 7 Dec 2020 05:51:53 GMT
• Title: Differentiable Augmentation for Data-Efficient GAN Training
• Authors: Shengyu Zhao, Zhijian Liu, Ji Lin, Jun-Yan Zhu, Song Han
• Abstract summary: We propose DiffAugment, a simple method that improves the data efficiency of GANs by imposing various types of differentiable augmentations on both real and fake samples. Our method can generate high-fidelity images using only 100 images without pre-training, while being on par with existing transfer learning algorithms.
• Score: 48.920992130257595
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The performance of generative adversarial networks (GANs) heavily deteriorates given a limited amount of training data. This is mainly because the discriminator is memorizing the exact training set. To combat it, we propose Differentiable Augmentation (DiffAugment), a simple method that improves the data efficiency of GANs by imposing various types of differentiable augmentations on both real and fake samples. Previous attempts to directly augment the training data manipulate the distribution of real images, yielding little benefit; DiffAugment enables us to adopt the differentiable augmentation for the generated samples, effectively stabilizes training, and leads to better convergence. Experiments demonstrate consistent gains of our method over a variety of GAN architectures and loss functions for both unconditional and class-conditional generation. With DiffAugment, we achieve a state-of-the-art FID of 6.80 with an IS of 100.8 on ImageNet 128x128 and 2-4x reductions of FID given 1,000 images on FFHQ and LSUN. Furthermore, with only 20% training data, we can match the top performance on CIFAR-10 and CIFAR-100. Finally, our method can generate high-fidelity images using only 100 images without pre-training, while being on par with existing transfer learning algorithms. Code is available at https://github.com/mit-han-lab/data-efficient-gans.

Related papers

• ACT-Diffusion: Efficient Adversarial Consistency Training for One-step Diffusion Models [59.90959789767886]
  We show that optimizing consistency training loss minimizes the Wasserstein distance between target and generated distributions. By incorporating a discriminator into the consistency training framework, our method achieves improved FID scores on CIFAR10 and ImageNet 64$times$64 and LSUN Cat 256$times$256 datasets.
  arXiv  Detail & Related papers  (2023-11-23T16:49:06Z)
• Improving Visual Representation Learning through Perceptual Understanding [0.0]
  We present an extension to masked autoencoders (MAE) which improves on the representations learnt by the model by explicitly encouraging the learning of higher scene-level features. We achieve 78.1% top-1 accuracy linear probing on ImageNet-1K and up to 88.1% when fine-tuning, with similar results for other downstream tasks.
  arXiv  Detail & Related papers  (2022-12-30T00:59:46Z)
• ScoreMix: A Scalable Augmentation Strategy for Training GANs with Limited Data [93.06336507035486]
  Generative Adversarial Networks (GANs) typically suffer from overfitting when limited training data is available. We present ScoreMix, a novel and scalable data augmentation approach for various image synthesis tasks.
  arXiv  Detail & Related papers  (2022-10-27T02:55:15Z)
• Augmentation-Aware Self-Supervision for Data-Efficient GAN Training [68.81471633374393]
  Training generative adversarial networks (GANs) with limited data is challenging because the discriminator is prone to overfitting. We propose a novel augmentation-aware self-supervised discriminator that predicts the augmentation parameter of the augmented data. We compare our method with state-of-the-art (SOTA) methods using the class-conditional BigGAN and unconditional StyleGAN2 architectures.
  arXiv  Detail & Related papers  (2022-05-31T10:35:55Z)
• Data-Efficient Instance Generation from Instance Discrimination [40.71055888512495]
  We propose a data-efficient Instance Generation (InsGen) method based on instance discrimination. In this work, we propose a data-efficient Instance Generation (InsGen) method based on instance discrimination.
  arXiv  Detail & Related papers  (2021-06-08T17:52:59Z)
• Ultra-Data-Efficient GAN Training: Drawing A Lottery Ticket First, Then Training It Toughly [114.81028176850404]
  Training generative adversarial networks (GANs) with limited data generally results in deteriorated performance and collapsed models. We decompose the data-hungry GAN training into two sequential sub-problems. Such a coordinated framework enables us to focus on lower-complexity and more data-efficient sub-problems.
  arXiv  Detail & Related papers  (2021-02-28T05:20:29Z)
• Dataset Condensation with Differentiable Siamese Augmentation [30.571335208276246]
  We focus on condensing large training sets into significantly smaller synthetic sets which can be used to train deep neural networks. We propose Differentiable Siamese Augmentation that enables effective use of data augmentation to synthesize more informative synthetic images. We show with only less than 1% data that our method achieves 99.6%, 94.9%, 88.5%, 71.5% relative performance on MNIST, FashionMNIST, SVHN, CIFAR10 respectively.
  arXiv  Detail & Related papers  (2021-02-16T16:32:21Z)
• Does Data Augmentation Benefit from Split BatchNorms [29.134017115737507]
  State-of-the-art data augmentation strongly distorts training images, leading to a disparity between examples seen during training and inference. We propose an auxiliary BatchNorm for the potentially out-of-distribution, strongly augmented images. We find that this method significantly improves the performance of common image classification benchmarks such as CIFAR-10, CIFAR-100, and ImageNet.
  arXiv  Detail & Related papers  (2020-10-15T15:00:43Z)
• Syn2Real Transfer Learning for Image Deraining using Gaussian Processes [92.15895515035795]
  CNN-based methods for image deraining have achieved excellent performance in terms of reconstruction error as well as visual quality. Due to challenges in obtaining real world fully-labeled image deraining datasets, existing methods are trained only on synthetically generated data. We propose a Gaussian Process-based semi-supervised learning framework which enables the network in learning to derain using synthetic dataset.
  arXiv  Detail & Related papers  (2020-06-10T00:33:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.