HANDS

HANDS数据集由东北大学创建，专注于模拟人类抓握意图推理，适用于假肢手研究。该数据集包含413个实例，涵盖多种日常物品，通过眼视角和手视角图像捕捉不同物品的不同角度。数据集创建过程中，使用了眼视角和手视角摄像头同步捕捉图像和视频，以及EMG和IMU传感器数据。HANDS数据集的应用领域主要集中在通过视觉和生理信号提高假肢手的感知能力和控制精度，解决假肢手在实际操作中的灵活性和准确性问题。

The HANDS dataset was developed by Northeastern University (China), focusing on simulating human grasping intention reasoning and tailored for prosthetic hand research. This dataset contains 413 instances covering a wide range of daily objects, capturing diverse viewpoints of various objects via eye-view and hand-view images. During the dataset construction process, eye-view and hand-view cameras were used to synchronously collect images, videos, as well as EMG and IMU sensor data. The primary application scope of the HANDS dataset centers on enhancing the perception capability and control accuracy of prosthetic hands through visual and physiological signals, aiming to address the limitations of flexibility and accuracy of prosthetic hands in real-world operational scenarios.