Multimodal Industrial Activity Monitoring (MIAM)
收藏资源简介:
MIAM数据集是一个多模态工业活动监控数据集,由印度科学与创新研究院中央电子工程研究所创建,旨在捕捉工业环境中的操作员行为。该数据集包含22个会话,总计290分钟的多视角RGB、深度和惯性测量单元(IMU)数据,详细标注了任务执行和操作员行为。数据集的内容涵盖了工业装配和拆卸任务,数据来源包括多视角摄像头和IMU传感器。数据集的创建过程包括在不受控环境中记录操作员的行为,并通过多模态数据融合来提升对操作员参与度的预测。该数据集的应用领域主要集中在工业环境中的人机协作研究,旨在通过多模态数据提升操作员行为监控和任务执行的效率。
The MIAM dataset is a multimodal industrial activity monitoring dataset developed by the Central Electronics Engineering Research Institute under the Academy of Scientific and Innovative Research (AcSIR), India, with the objective of capturing operator behaviors in industrial environments. It comprises 22 sessions, totaling 290 minutes of multi-view RGB, depth, and Inertial Measurement Unit (IMU) data, with comprehensive annotations for task execution and operator behaviors. The dataset covers industrial assembly and disassembly tasks, with data sourced from multi-view cameras and IMU sensors. The dataset creation workflow entails recording operator behaviors in uncontrolled environments, and utilizing multimodal data fusion to enhance predictions of operator engagement. Its primary application domains lie in human-robot collaboration research in industrial settings, aiming to improve the efficiency of operator behavior monitoring and task execution through multimodal data.

- 1A Multimodal Dataset for Enhancing Industrial Task Monitoring and Engagement Prediction印度科学与创新研究院中央电子工程研究所 · 2025年



