HackerPapers

arXiv:1912.04958cs.CV

Analyzing and Improving the Image Quality of StyleGAN

Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, Timo Aila

Abstract

The style-based GAN architecture (StyleGAN) yields state-of-the-art results in data-driven unconditional generative image modeling. We expose and analyze several of its characteristic artifacts, and propose changes in both model architecture and training methods to address them. In particular, we redesign the generator normalization, revisit progressive growing, and regularize the generator to encourage good conditioning in the mapping from latent codes to images. In addition to improving image quality, this path length regularizer yields the additional benefit that the generator becomes significantly easier to invert. This makes it possible to reliably attribute a generated image to a particular network. We furthermore visualize how well the generator utilizes its output resolution, and identify a capacity problem, motivating us to train larger models for additional quality improvements. Overall, our improved model redefines the state of the art in unconditional image modeling, both in terms of existing distribution quality metrics as well as perceived image quality.

arXiv comments
Journal reference
DOI
Subjects
cs.CVComputer Vision and Pattern Recognition cs.LGMachine Learning cs.NENeural and Evolutionary Computing eess.IVImage and Video Processing stat.MLMachine Learning (Statistics)

On Hacker News

Linked by 6 identifiable commenters in 7 comments across 7 threads, from Feb 2020 to Oct 2024. Found in 1 HN submission from Dec 2019 to Dec 2019. Open submission discussions.

Comments linking this paper, per year.

Discussed in comments7 threads

  1. Ask HN: What Are You Working On? (October 2024) 1 comment

    • I've been working a lot on my blog, which entails training and implementing neural networks. My next two posts are going to be about: - https://arxiv.org/pdf/1912.04958 - StyleGAN2 - https://arxiv.org/pdf/1807.09341 - Causal InfoGAN, written in TinyGrad https://ym2132.github.io

      Two_handsView on HNParent context

  2. Show HN: This Food Does Not Exist 1 comment

    • > The denoising part of a denoising autoencoder refers to the noise applied to its input Agree, it converts a noisy image to a denoised image. But the odd thing is, when you put a noisy image into a StyleGAN2 encoder, you get latents which the decoder will turn into a de-noised…

      fxtentacleView on HNParent context

  3. Alias-Free GAN 1 comment

    • To be honest I don't think the StyleGAN papers are benchmark chasing. If you read StyleGAN[0], StyleGAN2[1], StyleGAN2-ADA[2], and this paper there is a clear story. They call out the mistakes in the previous papers and resolve them. The papers themselves even admit to where…

      godelskiView on HNParent context

  4. Accounts with GAN Faces Attack Belgium over 5G Restrictions 1 comment

    • The StyleGAN2 paper[0] actually addresses some of the biggest clues. They didn't completely solve everything but symmetry is one of the biggest indicators. Specifically ears and eyes. But also really pay attention to the teeth. If you go to ThisPersonDoesNotExist[1] you'll see…

      godelskiView on HNParent context

  5. Ask HN: What's the best paper you've read in 2020? 1 comment
  6. This X Does Not Exist (2019) 1 comment

    • Seminal papers in this area AFAIK are StyleGAN1 and StyleGAN2: https://github.com/NVlabs/stylegan2 https://arxiv.org/abs/1912.04958

      FL33TW00DView on HNParent context

  7. This person does not exist 1 comment

    • Seems like it has been vastly improved since then [1]. The results are much better than when originally posted in HN. [1] - https://arxiv.org/abs/1912.04958

      dtakumaView on HNParent context

Submitted to Hacker News

Connected through shared HN threads, not inferred similarity or a recommendation.