eWEAR: Bringing greater reality to virtual reality
Meeting Reports
|
Jun 20, 2022
Exactly why that is has a lot to do with bandwidth—all that high resolution imagery and rapid framerates chew up data, fast. The computers can’t keep up. In a new paper, researchers at Stanford University recently explored a clever solution to the problem by mimicking the eye’s own mechanisms for sight.
The researchers built an innovative prototype of a “foveated” compression system that works in conjunction with gaze tracking. The fovea is the ophthalmological term for the area of the retina with greatest visual acuity.
This combination of eye-tracking, compression, and ultra-low latency is, the team believes, a first in Internet-based VR video systems. It is based on a client-and-server model, in which the client is the virtual reality headset. The client tracks and reports the viewer’s current gaze position for every frame, allowing the server to encode the next frame foveated on the viewer’s precise gaze position.
Exactly why that is has a lot to do with bandwidth—all that high resolution imagery and rapid framerates chew up data, fast. The computers can’t keep up. In a new paper, researchers at Stanford University recently explored a clever solution to the problem by mimicking the eye’s own mechanisms for sight.
The researchers built an innovative prototype of a “foveated” compression system that works in conjunction with gaze tracking. The fovea is the ophthalmological term for the area of the retina with greatest visual acuity.
This combination of eye-tracking, compression, and ultra-low latency is, the team believes, a first in Internet-based VR video systems. It is based on a client-and-server model, in which the client is the virtual reality headset. The client tracks and reports the viewer’s current gaze position for every frame, allowing the server to encode the next frame foveated on the viewer’s precise gaze position.
