Eric Wallace on Twitter: "Models such as Stable Diffusion are trained on copyrighted, trademarked, private, and sensitive images. Yet, our new paper shows that diffusion models memorize images from their training data and emit them at generation time. Paper: https://t.co/LQuTtAskJ9 👇[1/9] https://t.co/ieVqkOnnoX" / Twitter - discu.eu

Hacker News

Stable Diffusion found to emit some images from training at generation time https://twitter.com/Eric_Wallace_/status/1620449934863642624 6 comments 1/2/2023

Reddit

[R] Extracting Training Data from Diffusion Models https://twitter.com/eric_wallace_/status/1620449934863642624?s=46&t=GVukPDI7944N8-waYE5qcw 78 comments 1/2/2023 machinelearning

Linking pages

Paper: Stable Diffusion “memorizes” some images, sparking privacy concerns | Ars Technica https://arstechnica.com/information-technology/2023/02/researchers-extract-training-images-from-stable-diffusion-but-its-difficult/ 60 comments

Would you like to stay up to date with Computer science? Checkout Computer science Weekly.

Related searches:

Search whole site: site:twitter.com

Search title: Eric Wallace on Twitter: "Models such as Stable Diffusion are trained on copyrighted, trademarked, private, and sensitive images. Yet, our new paper shows that diffusion models memorize images from their training data and emit them at generation time. Paper: https://t.co/LQuTtAskJ9 👇[1/9] https://t.co/ieVqkOnnoX" / Twitter

See how to search.

Submit link to: