M2S Laboratory (Movement, Sports & Health), University Rennes 2, ENS Rennes, 35170 Bruz, France.
MIMETIC-Analysis-Synthesis Approach for Virtual Human Simulation, INRIA Rennes Bretagne Atlantique, Campus de Beaulieu, 263 Av. Général Leclerc, 35042 Rennes, France.
Sensors (Basel). 2022 Aug 3;22(15):5786. doi: 10.3390/s22155786.
This study presents a deep learning model devoted to the analysis of swimming using a single Inertial Measurement Unit (IMU) attached to the sacrum. Gyroscope and accelerometer data were collected from 35 swimmers with various expertise levels during a protocol including the four swimming techniques. The proposed methodology took high inter- and intra-swimmer variability into account and was set up for the purpose of predicting eight swimming classes (the four swimming techniques, rest, wallpush, underwater, and turns) at four swimming velocities ranging from low to maximal. The overall F1-score of classification reached 0.96 with a temporal precision of 0.02 s. Lap times were directly computed from the classifier thanks to a high temporal precision and validated against a video gold standard. The mean absolute percentage error (MAPE) for this model against the video was 1.15%, 1%, and 4.07%, respectively, for starting lap times, middle lap times, and ending lap times. This model is a first step toward a powerful training assistant able to analyze swimmers with various levels of expertise in the context of in situ training monitoring.
本研究提出了一种深度学习模型，用于通过附着在骶骨上的单个惯性测量单元（IMU）分析游泳。在包括四种游泳技术的方案中，从 35 名不同专业水平的游泳者中收集了陀螺仪和加速度计数据。所提出的方法考虑了高度的个体间和个体内变异性，并旨在预测八个游泳类别（四种游泳技术、休息、壁推、水下和转身），游泳速度范围从低到最大。分类的整体 F1 分数达到 0.96，时间精度为 0.02 秒。得益于高时间精度，泳道时间可以直接从分类器中计算出来，并通过视频黄金标准进行验证。与视频相比，该模型的平均绝对百分比误差（MAPE）分别为 1.15%、1%和 4.07%，用于起始泳道时间、中间泳道时间和结束泳道时间。该模型是朝着能够在原位训练监测的背景下分析具有不同专业水平的游泳者的强大训练助手迈出的第一步。
