TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

A Style-Based Generator Architecture for Generative Adversarial Networks [pdf]

78 pointsby cgtyoderover 6 years ago

4 comments

modelessover 6 years ago
Video: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;kSLJriaOumA?t=26" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;kSLJriaOumA?t=26</a><p>The quality and diversity of these images is incredible.
评论 #18677501 未加载
smusamashahover 6 years ago
I believe that best of these generated faces will have a 90% match in training dataset.<p>If you go through the generated faces you can see all of them have different background. These are not generated from scratch, only picked from memory to match the requirements.<p>Each face has totally different hair and hairstyle.<p>Have the published the training dataset?
treebroover 6 years ago
This is just phenomenal. Can see this being a a fairly disruptive force in the media industry.<p>Also, sock puppet factories could use this to create endless numbers of fake personas for social media astroturfing.
评论 #18677760 未加载
distant_hatover 6 years ago
The improvements in GANs from 2014 are amazing. From coarse 32x32 pixel images we have gotten to 1024x1024 images that can fool most humans.