smartwatch inertial data dataset
收藏资源简介:
本研究创建了一个包含10名参与者在半控制条件下进行饮食的智能手表惯性数据数据集,共记录了342次咬合事件。数据集包含了智能手表(华为Watch 2) worn on the dominant wrist收集的同步加速度计和陀螺仪信号,以及通过蓝牙启用的小菜盘秤测量的地面真实咬合重量。该数据集用于评估提出的咬合重量估计方法,该方法结合了从微运动分类模型中提取的行为特征和惯性信号的统计特征,以估计每次咬合的重量。
This study created a smartwatch-based inertial dataset for eating tasks conducted by 10 participants under semi-controlled conditions, with a total of 342 masticatory events recorded. The dataset includes synchronized accelerometer and gyroscope signals collected by a Huawei Watch 2 worn on the participants' dominant wrists, alongside ground-truth masticatory weight measurements acquired via a Bluetooth-enabled small plate scale. This dataset is employed to evaluate the proposed masticatory weight estimation method, which combines behavioral features extracted from a micro-movement classification model and statistical features of inertial signals to estimate the weight of each single chewing event.




