🤖 AI Summary
To address the computational bottleneck in rendering wall-mounted displays with printer-level ultra-high resolution (e.g., hundreds of millions of pixels) at hundreds of frames per second, this paper proposes a time-adaptive sampling rendering paradigm. Instead of relying on fixed frame rates and static pixel allocation, the method dynamically schedules pixel update regions and sampling instants based on content change, coupled with a lightweight driving algorithm to enable demand-driven rendering. Experiments demonstrate that, while preserving visual quality, the approach reduces pixel-level computation by over 90% compared to conventional frame-based rendering and supports real-time refresh rates of 120–500 Hz. The core innovation lies in jointly modeling spatial and temporal redundancy, enabling, for the first time, an efficient, low-latency, and scalable refresh architecture for ultra-large-scale pixel arrays. This work establishes a novel pathway toward billion-pixel display systems.
📝 Abstract
Could the answer be to compute fewer pixels? Renderers that break traditional framed patterns and opt for temporally adaptive sampling might be the key to printer-resolution wall displays that update hundreds of times per second.