Efficient Radius Search for Adaptive Foveal Sizing Mechanism in Collaborative Foveated Rendering Framework

Yufei Yang,Chenhao Xie,Liansheng Liu,Philip H. W. Leong,Shuaiwen Leon Song
DOI: https://doi.org/10.1109/tmc.2023.3277577
IF: 6.075
2024-01-01
IEEE Transactions on Mobile Computing
Abstract:Collaborative Foveated Rendering (CFR) is the latest collaborative rendering framework proposed to enable high frame rate VR applications on mobile devices. Compared with the strategies adopted in conventional collaborative rendering, the pixel-based Adaptive Foveal Sizing (AFS) mechanism in CFR offers a more flexible and intelligent workload trade-off by predicting the radius. However, the performance of the AFS mechanism in actual deployment depends on its adaptability to two factors, including the Sudden Environmental Variations (SEV) and the Random Discrete Latency (RDL). Guaranteeing the performance of the AFS mechanism by adapting to these two factors is of great significance to guaranteeing users' immersive experience. This paper identifies the existence of the SEV and RDL phenomenon in the AFS mechanism for the first time, and contributes the first method that offers the effective and real-time AFS mechanism implementation for the practical deployment, namely the Efficient Radius Search (ERS). The ERS method efficiently searches the largest radius online that controls the rendering workload within the foveated layer just below the offline baked threshold, thereby achieving the immediate response to SEV and reducing the oscillating frame rendering latency led by RDL. Through the experiments on 3 VR applications and 4 mobile devices, the resulting 2.44× to 9.07× higher frame rate precision compared with the state-of-the-art method demonstrate the superiority of the ERS method.
What problem does this paper attempt to address?