Asynchronous temporal mapping for high-dynamic-range video and hardware-level privacy with event cameras
摘要
Conventional imaging systems struggle to simultaneously achieve high-dynamic-range imaging under extreme illumination and ensure that visual data remain confidential from the moment of acquisition. Here we present Asynchronous Temporal Mapping, a framework that encodes high-dynamic-range visual information into a chaotic, asynchronous event stream through pixel-wise optical modulation. The framework addresses the bandwidth bottleneck of prior synchronous temporal-mapping approaches, enabling high-fidelity video capture with event cameras. By embedding the asynchronous principle of event sensing into a computational optical pipeline, it introduces an imaging framework characterized by high-dynamic-range video capture and inherent hardware-level protection.