This was definitely Adobe’s week with a huge series of product announcements including new Lightroom CC and Lightroom Classic versions, Photoshop and Premiere Pro 2018 as well as many other new and updated products.
But Adobe also had its yearly Adobe MAX convention where, amongst other things, it also showcased some of the very interesting technologies it is currently developing (many of them do indeed end up in future products by the company).
We shared with you several past Adobe Max sneak peek videos but it seems that this year a lot of what Adobe is working on has to do with the power of artificial intelligence and although we have already seen a little bit of what Adobe can do with AI in the new Lightroom CC, this is actually just the very tip of the iceberg compared to what is really going on in the labs of the company.
Above you can see one of the most exciting technologies showed by Adobe this year on Adobe Max called Project Cloak which can help you hide objects in a video in more or less the same way you can use content aware in Photoshop to remove a power line or any other distracting object that you don’t want.
As for as we know Project Cloak doesn’t use any AI (although maybe it will in the future – it can fill up all sorts of objects and it might need to know how to do that based on large databases of objects and patterns). What does use AI is a different project called DeepFill.
Project DeepFill uses a large database of images to create better fill options (you know how you hate Photoshop making those horrible fill results – this might be a thing of the past in the not too distant future).
Project DeepFill demo
The last project is true AI in its core – it looks a bit similar but what it does is allow you to replace elements in any picture by selecting them and than replacing them with smart content aware based on AI generated results and letting you choose from those – it is quite amazing to see how this works (we have to say though that at some point copyright will start to get into the picture here – especially if you are basically using a large portion of a completely different picture from Adobe’s database).
A peek into project SceneStich – AI driven object replacement