Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
Concept
Foveated rendering
Summary
Foveated rendering is a rendering technique which uses an eye tracker integrated with a virtual reality headset to reduce the rendering workload by greatly reducing the image quality in the peripheral vision (outside of the zone gazed by the fovea).
A less sophisticated variant called fixed foveated rendering doesn't utilise eye tracking and instead assumes a fixed focal point.
Research into foveated rendering dates back at least to 1991.
At Tech Crunch Disrupt SF 2014, Fove unveiled a headset featuring foveated rendering. This was followed by a successful kickstarter in May 2015.
At CES 2016, SensoMotoric Instruments (SMI) demoed a new 250 Hz eye tracking system and a working foveated rendering solution. It resulted from a partnership with camera sensor manufacturer Omnivision who provided the camera hardware for the new system.
In July 2016, Nvidia demonstrated during SIGGRAPH a new method of foveated rendering claimed to be invisible to users.
In February 2017, Qualcomm announced their Snapdragon 835 Virtual Reality Development Kit (VRDK) which includes foveated rendering support called Adreno Foveation.
During CES 2019 on January 7 HTC announced an upcoming virtual reality headset called Vive Pro Eye featuring eye-tracking and support for foveated rendering.
In December 2019, Facebook's Oculus Quest SDK gave developers access to dynamic fixed foveated rendering, allowing the variation in level of detail to be changed on the fly via an API.
On January 4, 2022, Sony announced that their follow-up to PlayStation VR will include eye tracking and foveated rendering.
On June 5, 2023, Apple announced that the Apple Vision Pro extended reality headset includes dynamic foveated rendering.
According to chief scientist Michael Abrash at Oculus, utilising foveated rendering in conjunction with sparse rendering and deep learning image reconstruction has the potential to require an order of magnitude fewer pixels to be rendered in comparison to a full image. Later, these results have been demonstrated and published.
Official source
About this result
This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.