Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces STAR-NT, a spatiotemporal acceleration framework designed to enhance the efficiency of neural order-independent transparency rendering by leveraging spatial and temporal coherence. By employing adaptive quadtree-based screen-space subdivision and depth-based reprojection, STAR-NT significantly reduces the computational overhead associated with geometry passes and network input generation, particularly on mobile and legacy hardware. The key result is a marked reduction in rendering costs while maintaining high visual quality, making real-time rendering more accessible across various platforms.
Achieving high-quality transparency rendering on legacy hardware is now feasible with STAR-NT's innovative spatiotemporal optimizations.
Neural order-independent transparency delivers high-quality rendering of overlapping transparent surfaces, but its geometry passes and network input generation remain costly, particularly on mobile and legacy hardware. We present a spatiotemporal acceleration framework that exploits spatial and temporal coherence to reduce this overhead while preserving visual quality. Spatially, we use adaptive quadtree-based screen-space subdivision to scale geometry pass resolution according to local color variance. Temporally, selected frames reuse the previous transparency result through depth-based reprojection instead of full rendering. Together, these optimizations reduce rendering cost and integrate efficiently into existing real-time rendering pipelines.