JerusML Hands-On sessions #1 - Demystifying the text to image magic

Event: fb://event/539462581330790

Details
This hands-on session will introduce you to the latest breakthroughs in image generation, including OpenAI's Dall-E2, Stable Diffusion, and Imagen. You'll learn how these models work and understand their key innovations, and get the opportunity to train a mini diffusion model and edit images using Stable Diffusion. The session will be conducted using Colab notebooks, so you'll need to have a laptop with you.
The instructor for this session will be Naomi Ken Korem, a Computer Vision Researcher at Lightricks. With her guidance, you'll get hands-on experience with the latest developments in image generation.
Before the hands-on session, it's important to take some preparationsteps. To get started, you'll need to create an account on HuggingFace and get a security token that will enable you to download stable diffusion models. This token is essential for accessing the models we'll be working with during the session. You can find instructions for creating an account andgetting a security token in the HuggingFace documentation at https://huggingface.co/docs/hub/security-tokens.
In addition to downloading the models, you may also want to prepare some images that you'd like to work with during the editing portion of the session. You may also want to generate binary masks of the areas in the image that you'd like to edit. If you don't have access to a photo editing tool like Photoshop, you can use an online tool like Photopea (https://http://www.photopea.com/). On Photopea, you can drag anddrop an image, add layers, and draw on the layers using different brushes. Take a look at the site and try generating some masks to get familiar with the tools you'll be using during the session.
Please note that there isa limited number of spots available for this hands-on session, and you can sign up by filling out an online form. The JerusML team will return an answer to those who get accepted.