This is the authors' abstract. We don't add key points for this paper.
Light-field displays (LFDs) fundamentally suffer from low spatial resolution, constrained by the pixel count of available microdisplays. Field-sequential color liquid crystal displays (FSC-LCDs) offer a promising solution by temporally mixing colors without color filters, thereby enabling a threefold increase in resolution. However, deploying FSC-LCDs for LFD applications remains challenging due to panel fabrication hurdles and the computational bottlenecks associated with rendering elemental image arrays (EIAs) and mitigating color breakup (CBU) artifacts at video rates. To overcome these challenges, we first develop a 2.3K-by-2.3 K (1500 ppi) FSC-LCD featuring a high aperture ratio and fast response time. We then propose a unified, video-rate computational framework. This pipeline achieves high efficiency through two key innovations: (i) an EIA module that leverages a pre-computed voxel-pixel look-up table (LUT) to bypass real-time ray tracing, and (ii) an FSC module that utilizes a lightweight neural network to infer optimal backlight signals for CBU suppression and color fidelity. We validated this end-to-end architecture on an LFD prototype. While maintaining accurate focus cues, optical reconstructions yield an angular voxel density of approximately 18 pixels per degree (PPD) across a 59.8° field of view, surpassing contemporary LFDs in total resolution. Finally, a dynamic 3D image acquisition and processing pipeline starting from a commercial depth camera demonstrated the system’s computational efficiency, with the core EIA and FSC modules executing in just 32.7 ms on a CPU-GPU computing platform, establishing a scalable foundation for next-generation extended reality (XR) displays.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: