eStonefish-Scenes: A Sim-to-Real Validated and Robot-Centric Event-based Optical Flow Dataset for Underwater Vehicles
eStonefish-Scenes: 一种经过仿真到现实验证且以机器人为中心的基于事件的光学流数据集用于水下车辆
机构 * Robotics Research Institute (ViCOROB), Universitat de Girona (UdG), Spain 2 School of Engineering \& Physical Sciences, Heriot-Watt University, Edinburgh, UK
专题命中 机器人数据与评测 :robotics(abstract);navigation(abstract);分类 cs.CV
AI总结 eStonefish-Scenes是首个针对水下应用的基于事件的光学流合成数据集,结合了仿真生成和现实验证,提升了水下机器人视觉导航的可靠性。
Comments This revised version extends the original, which lacked real-world validation. We added a real-world data acquisition study using a DAVIS346 camera on a BlueROV2, a homography-based ground-truth optical flow method with per-pixel uncertainty estimation, and a sim-to-real evaluation using a ConvGRU network trained on synthetic data and tested on real underwater sequences. arXiv admin note: text overlap with arXiv:2412.09209