FR-GESTURE
收藏资源简介:
FR-GESTURE是由希腊多所高校联合创建的面向应急救援人机交互的RGBD手势数据集,包含12种专为地面无人车(UGV)设计的战术手势命令。数据集包含3312组同步采集的RGB-深度图像对,覆盖7种距离(1-7米)和3种环境场景(2室内+1室外),通过双Intel RealSense D415摄像头多视角采集以增强多样性。数据采集过程结合了消防员实战手势反馈进行优化,旨在解决复杂灾害场景下非接触式机器人控制的关键需求,为视觉手势识别算法在应急救援领域的应用提供基准资源。
FR-GESTURE is an RGBD gesture dataset tailored for human-robot interaction in emergency rescue operations, jointly developed by multiple universities across Greece. Comprising 12 tactical gesture commands specifically designed for Unmanned Ground Vehicles (UGVs), the dataset contains 3312 synchronized RGB-depth image pairs, covering 7 distance ranges (1–7 meters) and 3 environmental settings (2 indoor, 1 outdoor). To enhance data diversity, all samples were collected from multiple perspectives using dual Intel RealSense D415 cameras. The data collection workflow was optimized based on feedback from firefighters' actual operational gestures, targeting the critical demand for contactless robot control in complex disaster scenarios. This dataset serves as a benchmark resource to facilitate the application of visual gesture recognition algorithms in the emergency rescue domain.
- 1FR-GESTURE: An RGBD Dataset For Gesture-based Human-Robot Interaction In First Responder Operations雅典哈罗科皮奥大学·信息与远程通信系; 金斯顿大学·网络与数字媒体系; 西马其顿大学·电气与计算机工程系; 雅典娜研究中心·阿基米德 · 2026年



