目录
2. Preliminaries: GANs and diffusion-based generative models
3. Diffusion-GAN: Method and Theoretical Analysis
3.1 Instance noise injection via diffusion
3.4 Theoretical analysis with Examples
4.1 Comparison to state-of-the-art GANs
4.2 Effectiveness of Diffusion-GAN for domain-agnostic augmentation
4.3 Effectiveness of Diffusion-GAN for limited data
Diffusion-GAN: 将GAN与diffusion一起训练
Diffusion-GAN主要关注的是 stabilize GAN training 的问题。为了实现GAN的稳定训练,将实例噪声注入鉴别器的输入理论上可行但缺少实践的验证。本文介绍了Diffusion-GAN,它采用了高斯混合分布,在前向扩散链条中的所有扩散步骤中引入实例噪声。将观察到的或生成的数据中扩散出的混合物的随机样本作为输入送入鉴别器。生成器通过前向扩散链反向传播其梯度进行更新,其长度是自适应调整的,以控制每个训练步骤中允许的最大噪声与数据比率。
Abstract
Generative adversarial networks (GANs) are challenging to train stably, and a promising remedy of injecting instance noise into the discriminator input has not been very effective in practice. In this paper, we propose Diffusion-GAN, a novel GAN framework that leverages a forward diffusion chain to generate Gaussian-mixture distributed instance noise. Diffusion-GAN consists of three
components, including an adaptive diffusion process, a diffusion timestep-dependent discriminator, and a generator. Both the observed and generated data are diffused by the same adaptive diffusion process. At each diffusion timestep, there is a different noise-to-data ratio and the timestep-dependent discriminator learns to distinguish the diffused real data from the diffused generated data. The generator learns from the discriminator’s feedback by backpropagating through the forward diffusion chain, whose length is adaptively adjusted to balance the noise and data levels. We theoretically show that the discriminator’s timestep-dependent strategy gives consistent and helpful guidance to the generator, enabling it to match the true data distribution. We demonstrate the advantages of Diffusion-GAN over strong GAN baselines on various datasets, showing that it can produce more realistic images with higher stability and data efficiency than state-of-the-art GANs.
生成对抗网络(GANs)的稳定训练具有挑战性,而在鉴别器输入中注入实例噪声的解决方法在实践中还不是很有效。在本文中,我们提出了一种新的GAN框架,它利用一个前向扩散链来产生高斯混合分布的实例噪声。扩散-gan由三个组成部分组成,包括一个自适应扩散过程、一个与扩散时间步长相关的鉴别器和一个发生器。观测数据和生成的数据都通过相同的自适应扩散过程进行扩散。在每个扩散时间步长,有不同的噪声-数据比,时间步长相关的鉴别器学习区分扩散的真实数据和扩散生成的数据。生成器通过前向扩散链的反向传播从鉴别器的反馈中学习,该扩散链的长度被自适应地调整以平衡噪声和数据水平。我们从理论上证明,鉴别器的时间步长依赖策略为生成器提供了一致和有用的指导,使其能够匹配真实的数据分布。我们在不同的数据集上展示了扩散-gan相对于强GAN基线的优势,表明它可以比最先进的GAN产生更真实的图像,稳定性和更高的数据效率。
主要贡献:
1) We show both theoretically and empirically how the diffusion process can be utilized to provide a model- and domain-agnostic differentiable augmentation, enabling data-efficient and leaking-free stable GAN training.【稳定了GAN的训练】
2) Extensive experiments show that Diffusion-GAN boosts the stability and generation performance of strong baselines, including StyleGAN2 , Projected GAN , and InsGen , achieving state-of-the-art results in synthesizing photo-realistic images, as measured by both the Fréchet Inception Distance (FID) and Recall score.【diffusion提升了原始只有GAN组成的框架的性能,例如styleGAN2,Projected GAN】
1. Introduction
3. Diffusion-GAN: Method and Theoretical Analysis
为了构造扩散-gan,我们描述了如何通过扩散注入实例噪声,如何通过正向扩散过程的反向传播来训练生成器,以及如何自适应地调整扩散强度。我们进一步提供了一个玩具例子说明的理论分析。
Q:这篇是做diffusion和gan结合的,可以参照一下看看能不能做cyclegan的形式,同时也可以调研一下有没有人follow这篇论文做了类似cyclegan的事情
①能不能做cycleGAN
必须能
②有人做cycleGAN了吗
有了 cyclediffusion 2023ICCV
补充:图像生成
出处:图像生成任务
现在可以补上一个Cycle-Diffusion了
后话:拖了很久,最近家里事情不少,手里的一篇论文也在大修
终于抽出时间继续读啦 但是小的愚笨 就只能读这些了
今天看CSDN看到有人以周报的形式记录 觉得相当不错
打算从明天开始写周报 不这样零零碎碎的记录了