On Noise Injection in Generative Adversarial Networks

Feng, Ruili; Zhao, Deli; Zha, Zhengjun

doi:10.48550/arxiv.2006.05891

Cited by 4 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, KernelGAN-FKP delivers comparable performance in handling image noise. The underlying reason might be that noise injection in the generator of GAN can help to circumvent over-fitting [14].…”

Section: Visual Resultsmentioning

confidence: 99%

Flow-based Kernel Prior with Application to Blind Super-Resolution

Liang¹,

Zhang²,

Gu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Kernel estimation is generally one of the key problems for blind image super-resolution (SR). Recently, Double-DIP proposes to model the kernel via a network architecture prior, while KernelGAN employs the deep linear network and several regularization losses to constrain the kernel space. However, they fail to fully exploit the general SR kernel assumption that anisotropic Gaussian kernels are sufficient for image SR. To address this issue, this paper proposes a normalizing flow-based kernel prior (FKP) for kernel modeling. By learning an invertible mapping between the anisotropic Gaussian kernel distribution and a tractable latent distribution, FKP can be easily used to replace the kernel modeling modules of Double-DIP and Ker-nelGAN. Specifically, FKP optimizes the kernel in the latent space rather than the network parameter space, which allows it to generate reasonable kernel initialization, traverse the learned kernel manifold and improve the optimization stability. Extensive experiments on synthetic and realworld images demonstrate that the proposed FKP can significantly improve the kernel estimation accuracy with less parameters, runtime and memory usage, leading to stateof-the-art blind SR results.

show abstract

Section: Visual Resultsmentioning

confidence: 99%

Flow-based Kernel Prior with Application to Blind Super-Resolution

Liang¹,

Zhang²,

Gu³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…It is thought that in generative modeling, the generator's intrinsic dimensionality should ideally match that of the real image manifold [33,45]. While the latter is hard to calculate [14], it has been estimated to values as low as 20-50 [16,42]. Yet, training remains effective with our seemingly excessive overparameterization.…”

Section: Discussionmentioning

confidence: 99%

Overparameterization Improves StyleGAN Inversion

Poirier-Ginter¹,

Lessard²,

Smith³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Fix 2D noise injection Existing studies (Karras et al, 2019;Feng et al, 2020) have showed that injecting per-pixel noise can increase the model's capability of modeling stochastic variation (e.g. hairs, stubble).…”

Section: Nerf Path Regularizationmentioning

confidence: 99%

StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis

Gu,

Liu,

Wang

et al. 2021

Preprint

View full text Add to dashboard Cite

We propose StyleNeRF, a 3D-aware generative model for photo-realistic highresolution image synthesis with high multi-view consistency, which can be trained on unstructured 2D images. Existing approaches either cannot synthesize highresolution images with fine details or yield noticeable 3D-inconsistent artifacts. In addition, many of them lack control over style attributes and explicit 3D camera poses. StyleNeRF integrates the neural radiance field (NeRF) into a style-based generator to tackle the aforementioned challenges, i.e., improving rendering efficiency and 3D consistency for high-resolution image generation. We perform volume rendering only to produce a low-resolution feature map and progressively apply upsampling in 2D to address the first issue. To mitigate the inconsistencies caused by 2D upsampling, we propose multiple designs, including a better upsampler and a new regularization loss. With these designs, StyleNeRF can synthesize high-resolution images at interactive rates while preserving 3D consistency at high quality. StyleNeRF also enables control of camera poses and different levels of styles, which can generalize to unseen views. It also supports challenging tasks, including zoom-in and-out, style mixing, inversion, and semantic editing. 1

show abstract

On Noise Injection in Generative Adversarial Networks

Cited by 4 publications

References 18 publications

Flow-based Kernel Prior with Application to Blind Super-Resolution

Flow-based Kernel Prior with Application to Blind Super-Resolution

Overparameterization Improves StyleGAN Inversion

StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis

Contact Info

Product

Resources

About