With the advent of smart phones all we suddenly have a good camera in Pocket 24 hours a day. Nothing, we become amateur photographers and as magically emerged the selfies, nothing new, but that has become a true phenomenon that does not show any symptom of slowdown.
However, by very good to make the Chamber of our Android smartphone, and by many megapixels that hostel inside, take a photo of yourself may result in a dismal portrait. But Adobe wants to put an end to this problem and has announced that his research team, powered by Adobe Sensei, is exploring new ways of using artificial intelligence and machine learning to help improve our selfies.
Through the publication of a video on YouTube, Adobe has shown that it is working on this objective. For it has shown an application that can take a normal selfie and then modified the image so it appears that the photo has been taken from a greater distance. Thus, the technology and the tools included in the application can convert an image selfie which might be too embedded in the frame or twisted in one that looks more to a photo that has been taken not by one, but by someone else. Even you can convert a 2D elsewhere with a certain depth image (3D).
In addition, the application allow you to move a little head in the selfie image to get a better angle.
Other special effects that are shown in the video is the Automatic portrait masking: with just touching with the finger, the application deletes the background of the selfie; then can be replaced that Fund but with a more pleasant depth and effect of field image. This way it seems that the picture had been taken with a traditional camera.
Finally, the video demo posted by Adobe also shows how we could find a photo that you want to emulate in our selfie and automatically apply filters and other effects that another photo of our new.
Abode has not announced when it will launch, or if anything will, an application like that has shown in that video, but it is very likely that the company offers something very similar in the future that seems close.
I leave you with the video in question:
