Fitness Repetition Monitoring Using Dynamic Time Warping in Wearable Gym Tracking Devices
Annotatsiya
This is Monitoring of repetitions in fitness which is an important part of wearable gym trackers, and it enables a user to track their performance and progress with a high level of accuracy. Monitoring repetitions will be a feature to identify and count repetitions in different workouts in real-time and with actionable feedback. Using the most common techniques currently, repeated monitoring is frequently affected by many limitations, such as sensor noise, personal variations in subject movement during exercises, and dynamic non-linear time space distortions related to exercise and can preclude the replication accuracy and reliability. In order to surpass and outperform these shortcomings and suggest a novel monitoring framework, Monitoring Fitness Repetition - Dynamic Time Warping (MFR-DTW). The proposed approach involves DTW - the successful algorithm to estimate the closeness between two temporal sequences, over time and/or velocity to dynamically match the time series of the sensor data streams, of a wearable device. The experimental findings show that improvements are critical in major parameters which confirm that MFR-DTW is superior to the fixed-threshold (FT) and the single-output methods. In particular, MFR-DTW has better repetition counting accuracy of 92, better exercise recognition F1-score of 93.8, better robustness to variation of users of 90 and more accurate prediction of injury risk of 90.9. These advantages make MFR-DTW be more reliable and personalized in terms of real-time feedback when used in wearable devices, which can help the proper monitoring of exercise progress and injury prevention throughout time.