手势识别测试数据集

面向终身学习教学场景的手势识别测试研究，基于罗技C920摄像头采集，数据基于室内、室外场景，在手势指令发出人距离摄像头1m以内，构建包含15类手势在内的手势动作的测试视频共2224段，每段长度约10秒，每段视频包含一个手势的一次动作，包含抬手、做手势、放手的完整过程；共包含15类手势，其中8类为静态手势(点赞、握拳、V手势、比心、OK手势、摇滚、爱你、6手势)，7类为动态手势(左挥、右挥、上挥、下挥、食指顺时针旋转、食指逆时针旋转、手掌左右摆动)，数据量7.11G。该数据集支撑了专利《手势识别方法、装置、设备及存储介质》。
该数据集未来可用于开发和测试手势识别算法，特别是在复杂场景下的手势识别性能评估。研究人员可以利用该数据集验证模型在不同环境（如室内、室外）和不同手势类型（静态与动态）中的表现，从而提升手势识别系统的鲁棒性和准确性。此外，该数据集还可用于教育技术领域的研究，帮助开发智能教学交互系统，实现更自然的人机交互体验，提升在线教学的互动性和效率。
对于手势识别领域的研究，该数据集提供了多样化的手势动作视频，涵盖了静态和动态手势的完整过程。该数据集不仅有助于提高手势识别系统的性能，还能为教育技术的发展提供数据支持，具有重要的学术价值和实际应用意义。

A gesture recognition test study targeting lifelong learning teaching scenarios. Collected using a Logitech C920 webcam, the dataset was gathered in both indoor and outdoor environments, with the gesture performer standing within 1 meter of the camera. A total of 2224 test videos of gesture movements covering 15 gesture categories were produced. Each video is approximately 10 seconds long, contains one single gesture movement, and covers the complete process of raising the hand, performing the gesture, and lowering the hand. The 15 gesture categories consist of 8 static gestures: "thumbs up", "fist", "V-sign", "heart gesture", "OK sign", "rock gesture (devil horns)", "love you gesture", and "number 6 gesture"; and 7 dynamic gestures: "left swing", "right swing", "upward swing", "downward swing", "index finger clockwise rotation", "index finger counterclockwise rotation", and "left-right palm swing". The total size of the dataset is 7.11 GB. This dataset supports the Chinese patent titled "Gesture Recognition Method, Apparatus, Device and Storage Medium".

This dataset can be used for developing and testing gesture recognition algorithms, especially for evaluating gesture recognition performance in complex scenarios. Researchers can use this dataset to validate model performance across different environments (e.g., indoor and outdoor) and different gesture types (static and dynamic), thereby improving the robustness and accuracy of gesture recognition systems. Furthermore, this dataset can also be applied to research in educational technology, assisting in the development of intelligent teaching interaction systems to achieve more natural human-computer interaction experiences and enhancing the interactivity and efficiency of online teaching.

For research in the field of gesture recognition, this dataset provides diverse gesture movement videos covering the complete processes of both static and dynamic gestures. It not only helps improve the performance of gesture recognition systems but also provides data support for the development of educational technology, holding important academic value and practical application significance.

搜集汇总

背景与挑战

背景概述

该数据集面向终身学习教学场景，用于手势识别测试研究，包含通过摄像头采集的室内外场景下15类静态与动态手势的2224段视频。数据量约7.11G，支撑了相关专利，可用于开发和评估手势识别算法，并支持教育技术领域的人机交互研究。