Unsupervised explainable activity prediction in competitive Nordic Walking from experimental data

TL;DR

This work applies an online processing unsupervised clustering approach based on low-cost wearable inertial measurement units to achieve automatic explainability for predictions related to athletes’ activities, distinguishing between correct, incorrect, and cheating practices in Nordic Walking.

Abstract

Artificial Intelligence (AI) has found application in Human Activity Recognition (HAR) in competitive sports. To date, most Machine Learning (ML) approaches for HAR have relied on offline (batch) training, imposing higher computational and tagging burdens compared to online processing unsupervised approaches. Additionally, the decisions behind traditional ML predictors are opaque and require human interpretation. In this work, we apply an online processing unsupervised clustering approach based on low-cost wearable Inertial Measurement Units (IMUs). The outcomes generated by the system allow for the automatic expansion of limited tagging available (e.g., by referees) within those clusters, producing pertinent information for the explainable classification stage. Specifically, our work focuses on achieving automatic explainability for predictions related to athletes' activities, distinguishing between correct, incorrect, and cheating practices in Nordic Walking. The proposed solution achieved performance metrics of close to 100 % on average.

Figures (5)

• Figure 1: Nordic Walking practice assessment scheme.
• Figure 2: Arrangement of the imus. a) positions on body and poles, b) wrist imu, c) ankle imu, d) pole imus, e) imu size in cm.
• Figure 3: Accuracy curve for the nwgti data set.
• Figure 4: Cross entropy loss curve for the nwgti data set.
• Figure 5: Screenshot of explainability dashboard.