A lightweight foveation codec for VR​

Video details

At #SIGGRAPH2021, we presented a new foveated VR codec that performs a gaze-based image size reduction before video encoding. It decreases the amount of data to be transferred, which reduces bandwidth use and transfer latency. It does this while maintaining the quality of the area where the user is looking in the VR headset. The pre-video encoding compression levels of up to 6:1 are feasible. It is beneficial when it comes to consuming videos in VR or streaming content to VR headsets. We are working with NVIDIA Design and Visualization to explore the frontier of this technology and welcome more partners to join our exciting journey.