Quick Guide to MotionPortrait

Company Background

MotionPortrait spun out of Sony Kihara Research Center which, in addition to developing the graphic engine for the PlayStation, was also responsible for the development of Sony's cutting edge image processing technologies. When MotionPortrait was founded, it received ownership of many R&D outputs from Sony, and engineers skilled in image processing and CG technology transferred as well. We are responsible for the development of all of our technologies, and making use of our experience in R&D, we continuously improve them.

Core technologies

Face detection

Detect faces in photos -- recognize faces quickly and accurately, using machine learning technologies developed from massive sets of facial image data.

Our technology can detect multiple faces.

Additionally, it can detect human-like faces, such as cartoon or CG characters, and even dog and cat faces, to some degree.

face detection

Face analytics

Every face is composed of eyes, a nose, a mouth, and other parts, and no two faces are exactly the same.

Using our unique facial analysis algorithms, we can quantify various facial parameters.

Our technology can also analyze various attributes, such as gender, age, or emotion.

face analytics

Face animation

Precisely identify a face in a photo, and automatically synthesize a 3D model of it.
The model blinks and breathes, and the natural animation makes it seem almost alive.

What's more, it can be made to perform joyful, angry, sad, and other facial expressions; it can also be made to lip-sync with voice recordings.