What’s new at Adobe Max

Adobe Max is the biggest annual gathering of Adobe software suite users. During Adobe Max, reasearchers come and present their work during a special session: Sneaks. It’s a relaxed atmosphere and the audience has a kind attitude towards researchers and engineers who are a bit stressed by this vast audience. These studies are a kind of barometer of the technology progress that are or maybe one day will be integrated to the Adobe suite.

One of the first projects –All in– enables to extract a person from a photo and add it to another photo. What we are looking for here is a total automation of the process using Sensei, Adobe’s AI. It works pretty well on the examples they showed as far as trimming is concerned. It’s not that good when you look at the lighting though. Let’s keep an eye on that to see howAdobe wants to integrate that feature and the level of control available.

ImageTango is a project that makes it possible to texture a shape with the help of a photo. For the past few years, we have seen projects dealing with style recognition applied to a picture or a video, or tools which can describe an image and AI can create the described picture. In that case, you upload a shape and then or or several photos with the textures or the color display that you would like to have and you get variations around your shape / texture. It’s pretty impressive and I would love to see its behaviour with higher resolution images.

GoFigure, which tracks a filmed person, was also a quite impressive presentation. Control dots are created and you can assign them to an avatar for instance. It does not replace evolved tracking systems but I think this project will be useful for simple animation, motion design or maquette. This tool is already alsmost integrated to After Effects.

SweetTalk is an additional project : you get a sound file and it animates a drawing, a photo or an avatar according to the text that is read. It works pretty weel even thoutgh it lacks accuracy.

LightRight is most likely the project that concerns us most directly. The goal is to understand the volumes of a photo or a video to relight it. I admit results vary depending on the photos or videos that are shown. But it is also the interest of these previews: show us the remaining path before it can be used in production.

GlowStick project is presented within Adobe
Illustrator. It’s pretty smart and it will give illustrators great opportunities. The idea is to turn Illustrator shapes into lighting sources for our illustration. The other shapes block the light and cast shadows. That could be done “manually” using color gardient but it is so much faster with this tool.

The AboutFace project is a tool which makes it possible to check if an image has been altered. The tool gives us a probability rate for the image to be altered, the area of alteration and even gives a possibility to erase the alteration.