In January we wrote about an AI called Volume that ripped characters out of videos. The developers of that AI, Or Fleisher and Shirin Anlen, afresh adapted their website with an alternate AI agreement that converts any flat imageto 3D.

Credit: Volume

We’ve seen agnate AI before, but this one does more than facial reconstruction. We asked Fleisher, who you may also bethink as the architect of an AI that gives any webcam Kinect-like capabilities, if training an AI to catechumen any type of image presents a altered claiming than training one to catechumen article specific:

It is altered in some ways and agnate in others. While the mechanics of amalgam and training such deep Neural Networks have similarities, our mission to create a model that understands the human body and environments requires a lot of data… In the background, we are training a few models for altered tasks and developing a allocation method for anecdotic the ascendant elements in the image. Once we know what to expect, we are able to pick a model which has a more aesthetic compassionate of how this should look like in 3D. This is quite altered from other approaches that accrue a lot of ability into one single apparatus acquirements model. It is still work in progress, but we hope to keep on training the networks to access more ability appropriate to reconstruct a 3D human abstracts and environments from 2D images and videos.

Credit: Volume

It is important to say, we do not share any of the data users upload. We do about use it to appraise the apparatus acquirements models to get a better compassionate of what works and what doesn’t. Our goal with aperture our API to the public via our new website is also acquirements what are the tendencies, trends, and desires people have for such functionality. We also animate users to share their Volume reconstructions socially using the share icon at the top. We accept that by absorption the work we do internally to people online, we will be able to grow (with) this into an able and avant-garde tool.

Volume is being developed as an end-to-end band-aid for converting flat 2D images into 3D abstracts and environments. While the API is still in early development, it’s pretty easy to see this acceptable a go-to appliance suite for video editors and game developers. And, according to Fleisher, Volume will abide to get new features:

We are also attractive into more accepted areas such as image augmentation. We experimented and built an appliance called ReTouch, which enables the use of the depth admiration for retouching images in 2.5D, this was all open-sourced. With that being said, the purpose of the Volume API is to animate the assorted use of the same technology in altered fields and uses. By architecture Volume as a cloud, computer vision API we are able to accommodate the same after-effects we use for 3D about-face for people who aren’t necessarily using it for the same purpose.

Credit: Volume

You can check out the Volume API and catechumen your own images into 3D using AI at the team’s website.

Read next: Google Maps now lets businesses mark themselves as 'Women-Led'