Real-time Monitoring of Lower Limb Movement Resistance Based on Deep Learning

TL;DR

Experimental results demonstrate that MMTL-Net significantly outperforms existing models on the UCI Human Activity Recognition and Wireless Sensor Data Mining Activity Prediction datasets, achieving a lower Force Error Rate (FER) and higher Resistance Prediction Accuracy (RPA) than existing models.

Abstract

Real-time lower limb movement resistance monitoring is critical for various applications in clinical and sports settings, such as rehabilitation and athletic training. Current methods often face limitations in accuracy, computational efficiency, and generalizability, which hinder their practical implementation. To address these challenges, we propose a novel Mobile Multi-Task Learning Network (MMTL-Net) that integrates MobileNetV3 for efficient feature extraction and employs multi-task learning to simultaneously predict resistance levels and recognize activities. The advantages of MMTL-Net include enhanced accuracy, reduced latency, and improved computational efficiency, making it highly suitable for real-time applications. Experimental results demonstrate that MMTL-Net significantly outperforms existing models on the UCI Human Activity Recognition and Wireless Sensor Data Mining Activity Prediction datasets, achieving a lower Force Error Rate (FER) of 6.8% and a higher Resistance Prediction Accuracy (RPA) of 91.2%. Additionally, the model shows a Real-time Responsiveness (RTR) of 12 milliseconds and a Throughput (TP) of 33 frames per second. These findings underscore the model's robustness and effectiveness in diverse real-world scenarios. The proposed framework not only advances the state-of-the-art in resistance monitoring but also paves the way for more efficient and accurate systems in clinical and sports applications. In real-world settings, the practical implications of MMTL-Net include its potential to enhance patient outcomes in rehabilitation and improve athletic performance through precise, real-time monitoring and feedback.

Figures (6)

• Figure 1: Application of wearable technology in lower limb movement resistance monitoring and physical condition assessment. (A) Wearable systems for "smart" living and exercise management. (B) Signal flow in wearable devices for inferring triaxial motion velocity, acceleration, angular velocity, and attitude angles. (C) Integration of motion data into deep learning neural networks for identifying personal identity, movement state, and estimating motion speed in real-time.
• Figure 2: Schematic representation of the parameters critical for accurate lower limb movement resistance estimation. The left diagram shows the calculation of joint rotation angles ($\theta_{\text{hip}}$ and $\theta_{\text{knee}}$), while the right diagram depicts stride length ($l_1$, $l_2$) and associated angles ($\theta$ and $\varphi$), both essential for understanding the mechanical load on joints and assessing gait efficiency.
• Figure 3: The overall architecture and workflow of the Mobile Multi-Task Learning Network (MMTL-Net) designed for real-time lower limb movement resistance monitoring. The system includes the Data Input Module for receiving and preprocessing sensor data, the MobileNetV3 Feature Extraction Module for efficient feature capture, and the Multi-Task Learning (MTL) Module that simultaneously performs activity recognition and resistance estimation.
• Figure 4: Detailed architecture of the MobileNetV3 model used for feature extraction in MMTL-Net. The architecture includes depthwise separable convolutions, squeeze-and-excitation modules, and the swish activation function, all contributing to efficient and accurate processing of lower limb movement sensor data.
• Figure 5: Performance comparison of different methods. The top section shows Resistance Prediction Accuracy (RPA), and the bottom section shows Real-time Responsiveness (RTR) and Throughput (TP). This visual comparison highlights the superior performance of MMTL-Net across key metrics, reinforcing its robustness and effectiveness in real-time lower limb movement resistance monitoring.