AI + Adobe = Scene Stitch & Project Cloak


With AI technology advancing on almost a daily basis, it has now become easier than ever to edit photos and images. At the recently held Adobe MAX conference, the company showed off its latest prototype AI tools, developed by Adobe’s Sensei unit. This is a unit inside Adobe building a platform to harness the power of AI to improve Adobe’s products.

Say Hello to Adobe Scene Stitch

The new tools by Adobe are a bit more serious though. For example, Scene Stitch is an more advanced form of Content Aware Fill. If you have used Adobe products before, you would probably recognize this as being Adobe’s way of removing unwanted objects from images. But Content Aware Fill merely analyses the image to look for a suitable filler. Scene Stitch actually goes into Adobe’s stock photo library as a paint palette. So it will analyze the image being edited and then switch out similar scenes.

Adobe Scene Stitch
Adobe Scene Stitch in action at Adobe MAX
Image Credits: ICS Media

In addition, Adobe is also working on making Scene Stitch semantically aware. This means that it would not only identify the composition of the image, but also the content. So you can use it to swap out items such as Christmas tree with a coconut tree, or Lamborghini with a Tata Nano. This would also make it more challenging to identify edited images because it’s not just looking for repeated patches of pixels.

Editing videos became a lot easier with Project Cloak

Project Cloak works in a similar manner to Content Aware Fill. The only difference is that this is for video. Project Cloak allows you to separate a specific part of a clip such as an element that would disturb an otherwise perfect shot. Then you can simply remove that object with the press of a button. Previously, this would have to be carried out on a frame-by-frame basis. A time consuming task that required great skill.

While this seems to be a boon for content creators of both photo and video, it also means that because it’s easier, anyone can do it. For example, fake news can be accelerated because of these new tools, to create more realistic fake news. As the saying goes, “With great power, comes great responsibility”. Even if you say seeing is believing, how does one differentiate fact from fiction when tools like this can be used to blur the line between fact and fiction?

What are your thoughts on this? Love it? Hate it? Leave a comment below.


Please enter your comment!
Please enter your name here