Streaming Load and Rendering Optimization for Mobile VR or AR Art Exhibitions

Abstract

With the growing application of mobile VR or AR technologies in art exhibitions, immersive experiences demand higher efficiency in large-scale data streaming and stereoscopic rendering quality. However, mobile devices are constrained by limited computing power, storage capacity, and network bandwidth, making it challenging for traditional loading and rendering approaches to balance data transmission efficiency with user experience. Current research often relies on fixed-pattern viewpoint prediction based on historical data, which fails to adapt to the dynamic and personalized nature of user viewpoints in mobile environments. This mismatch leads to inefficient preloading, where loaded content does not align well with users’ actual focus areas. Furthermore, conventional stereoscopic rendering optimization strategies frequently overlook device performance variability and localized user attention, resulting in low rendering efficiency and excessive power consumption. To address these challenges, this study focuses on two core aspects: (1) viewpoint prediction for dynamic streaming preloading, which integrates user gaze patterns, attention distribution, and exhibition content structure to build a dynamic prediction model that enhances preloading relevance; and (2) viewpoint-driven stereoscopic rendering optimization, where high-detail rendering is applied to predicted focus regions while non-focus areas are simplified. This approach ensures visual quality in key scenes while reducing computational load. The outcomes of this study provide efficient technical solutions for mobile VR or AR art exhibitions, significantly improving streaming speed and rendering smoothness and promoting deeper integration between artistic presentation and digital technology.