Welcome to your

AI Weekly Digest

presented by Louis (What's AI) Bouchard

1️⃣ What is Stable Diffusion? (Latent Diffusion Models Explained)


What do all recent super powerful image models like DALLEImagen, or Midjourney have in common? Other than their high computing costs, huge training time, and shared hype, they are all based on the same mechanism: diffusion.

Diffusion models recently achieved state-of-the-art results for most image tasks including text-to-image with DALLE but many other image generation-related tasks too, like image inpainting, style transfer or image super-resolution. But how do they work? Read more...

Watch the video

2️⃣ (Sponspor) Superior ML demands Superior infrastructure

Qwak is a fully managed platform that unifies ML engineering and data operations - providing an agile infrastructure that enables the continuous productionization of ML models at scale.
Try out Qwak!

3️⃣ AI Ethics with Lauren

Stable diffusion is an incredible tool which has already demonstrated its power and versatility since its first use! Given Robin Rombach et al.’s work demonstrating the ability to use a mere image representation in place of an image and generate a full image based on it, the ethical application that came to mind for me was obfuscation.

Loosely defined, obfuscation is the concept of protecting one’s privacy and protesting the overuse of personally identifiable information with noise in various forms, which provides obfuscation. A great application of obfuscation is Fawkes, a tool from the University of Chicago Department of Computer Science co-led by Emily Wenger and Shawn Shan. The innovative tool adds noise to personal image to prevent facial recognition algorithms from properly identifying a person, misidentifying them in most cases and adding a layer of privacy protection. This obfuscates a person’s biometric data and respects their wish to remain anonymous, often when there is no option to do so.

If we combine these two incredible technologies, stable diffusion may counteract the efforts of Fawkes by reversing its obfuscation efforts, processing the noise-addled image to reverse engineer a person’s identity, revealing biometric data and violating the wish to remain anonymous. This is a specific application and is not meant to generalize about stable diffusion, but it is presented as an interesting use case that warrants more investigation and should be analyzed through different privacy lenses. On the flip side, we already have many positive use cases of stable diffusion, including some of the most hilarious and entertaining memes of recent years. So far the applications have been incredibly useful and positive, and I am excited to see many more use cases to come!

- AI Ethics segment by Lauren Keegan
Want to get into AI or improve your skills? Click here!
We are already at the end of this AI weekly digest! Thank you for thoroughly going through this iteration! I hope you enjoyed it. Feel free to follow my daily newsletter where I share 2 papers and news with a clear one-liner summary, and our newsletter at Towards AI, sharing the most exciting news, papers, articles, and memes weekly.

If you have suggestions, comments, or other thoughts, you can reach me by replying to this email or directly on Twitter or Linkedin. Don't hesitate to come to chat with more than 20'000 AI enthusiasts on Discord!
I hope the next week wipes away some of your stress and brings new opportunities, challenges and happiness. A happy new week from your friend!

If you would like to support my work financially, you can become a Patreon and receive a cool role in the discord server at the same time!

Share the knowledge and forward this email to a friend using this link:

- Louis Bouchard
Copyright © 2022 What's AI, All rights reserved.

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.

Email Marketing Powered by Mailchimp