OpenPack
收藏资源简介:
OpenPack是由大阪大学信息科学与技术研究生院创建的大型多模态工作活动数据集,专注于工业领域的包装工作识别。该数据集包含53.8小时的多模态传感器数据,涵盖加速数据、关键点、深度图像以及来自16个不同经验水平的包装工作参与者的物联网设备读数。OpenPack不仅规模庞大,还提供了丰富的元数据,如参与者的包装工作经验和物理特征,支持多种研究任务的设计。此外,它是首个包含物联网设备读数的大型数据集,旨在通过传感器和稀疏物联网数据实现复杂的动作识别,为普及计算社区提供挑战性任务。
OpenPack is a large-scale multimodal work activity dataset developed by the Graduate School of Information Science and Technology, Osaka University, focusing on industrial packaging work recognition. This dataset contains 53.8 hours of multimodal sensor data, including accelerometer data, human keypoints, depth images, and IoT device readings from 16 participants with varying levels of packaging work experience. Boasting a substantial scale, OpenPack also provides rich metadata such as participants' packaging work experience and physical characteristics, which facilitates the design of multiple research tasks. Furthermore, it is the first large-scale dataset to incorporate IoT device readings, aiming to enable complex action recognition using sensor and sparse IoT data, and offering challenging tasks for the pervasive computing community.

- 1OpenPack: A Large-scale Dataset for Recognizing Packaging Works in IoT-enabled Logistic Environments大阪大学信息科学与技术研究生院 · 2024年



