About Pro-Fi

On device hardware is finally pretty good and I think we should try to extract all the performance possible from the devices we own.

There is a lot more we can do in the future in terms of customizing what we watch and how we watch it. From simply optimizing the smoothness, sharpness and color, to going much deeper into 2D to 3D projection. Action cameras like Insta360 and live sports leave a lot to be desired currently. My bet here is that we can create smart hybrid local + remote intelligence that makes content really live up to the original immersive pitch that VR devices led with.

Pro-Fi started off as a Vision Pro project then quickly branched off to support all main Apple hardware, with explorations happening into Meta and beyond.

Research

We are actively developing our own internal ML processes and models for optimization of videos, anti-aliasing, and content understanding. We are looking to have as much as possible run local on-device.

What we’re researching

Content quality optimization

We’re exploring how to improve the quality of the videos you watch.

Conceptual detail reconstruction studyA sample grid beside a smooth contour illustrates research into video detail. This is a conceptual drawing, not a measured comparison.

Inline customization and correction

We’re looking at ways to customize and correct video as you watch.

Conceptual video adjustment studyA video frame with alignment guides and adjustment lines illustrates customization and correction while watching. It does not depict a shipped interface.

3D projection

We’re researching how to make fully convincing 2D to 3D projection without the typical artifacts.

Conceptual 2D to 3D projection studyRays connect a flat image plane to offset depth planes, illustrating the research direction rather than a demonstrated projection result.