ESVIO: Event-Based Stereo Visual Inertial Odometry

Event cameras that asynchronously output low-latency event streams provide great opportunities for state estimation under challenging situations. Despite event-based visual odometry having been extensively studied in recent years, most of them are based on the monocular, while few research stereo vision. In this letter, we present ESVIO, first visual-inertial odometry, which leverages complementary advantages streams, standard images, and inertial measurements. Our proposed pipeline includes ESIO (purely event-based) ESVIO (event with image-aided), achieves spatial temporal associations between consecutive streams. A well-design back-end tightly-coupled fused multi-sensor measurement to obtain robust estimation. We validate both have superior performance compared other image-based baseline methods public self-collected datasets. Furthermore, use our perform onboard quadrotor flights low-light environments. Autonomous driving data sequences real-world large-scale experiments also conducted demonstrate long-term effectiveness. highlight work is a real-time, accurate system aimed at