Position-Aware Multi-Sensor (PAMS)
收藏资源简介:
在本工作中,我们引入了一个新的数据集,称为位置感知多传感器(PAMS)。该数据集包含加速度计数据和陀螺仪数据。陀螺仪数据提高了活动识别方法的准确性,并使它们能够检测更广泛的活动。我们还考虑了用户信息。根据参与者的生物特征属性,生成单独的学习模型来分析他们的活动。我们专注于几个主要活动,包括坐、站、走、跑、上下楼梯和骑自行车。为了评估数据集,我们使用各种分类器,并将结果与WISDM进行比较。结果显示,使用上述分类器的所有活动的平均精度均超过88.5%。
In this work, we introduce a novel dataset termed Position-Aware Multi-Sensor (PAMS). This dataset encompasses accelerometer and gyroscope data. The gyroscope data enhances the accuracy of activity recognition methodologies and enables the detection of a broader spectrum of activities. We also take into account user information. Based on the biometric attributes of participants, individual learning models are generated to analyze their activities. Our focus is on several primary activities, including sitting, standing, walking, running, ascending and descending stairs, and cycling. To evaluate the dataset, we employ various classifiers and compare the results with those from WISDM. The findings indicate that the average accuracy for all activities using the aforementioned classifiers exceeds 88.5%.
数据集概述
数据集名称
Position-Aware Multi-Sensor (PAMS)
数据类型
- 加速度计数据
- 陀螺仪数据
数据用途
用于活动识别,包括以下活动:
- 坐
- 站立
- 行走
- 跑步
- 上下楼梯
- 骑自行车
数据特点
- 结合用户生物特征信息,为每个参与者生成单独的学习模型。
- 使用多种分类器进行评估,平均精度超过88.5%。
相关比较
与WISDM数据集进行比较。




