Landing Signalman gesture recognition dataset

These two gesture datasets are from the Computer & Graphics Journal submitted article: "Helicopter visual signaling simulation: Integrating VR and ML into a low-cost solution to optimize Brazilian Navy training".
Both files have 17 helicopter visual signals using two hands, and each gesture was performed 25 times. They were created using MiVRy – 3D Gesture Recognition at Unity environment. They contain the raw relative positions of each gesture sample, the neural network itself, and the metadata.

File “HVSS_gest_contr.dat” was recorded using VIVE(TM) controllers, for the virtual nighttime environment, and “HVSS_gest_tracker.dat” using VIVETM Tracker wrist-mounted, for the virtual daytime environment. The gestures are illustrated and explained according to "'Landing Signalman dataset.pdf".

这两组手势数据集源自《计算机与图形学》杂志提交的论文：“直升机视觉信号模拟：将虚拟现实与机器学习整合至低成本解决方案以优化巴西海军训练”。两组数据文件均包含17种使用双手执行的手势，每种手势重复了25次。这些数据通过在Unity环境中使用MiVRy – 3D手势识别技术创建。数据中包含了每个手势样本的原始相对位置、神经网络本身以及元数据。

文件“HVSS_gest_contr.dat”使用VIVE™控制器记录，针对虚拟夜间环境，而“HVSS_gest_tracker.dat”则使用VIVETM Tracker腕带式追踪器记录，针对虚拟白天环境。手势的演示和解释参照了“Landing Signalman dataset.pdf”。” }