Using deep neural networks to detect non-analytically defined expert event labels in canoe sprint force sensor signals

TL;DR

This paper investigates the use of CNNs and RNNs, particularly BGRUs, for automatic detection of paddle stroke events in canoe sprint force signals, aiming to reduce manual analysis and improve performance assessment.

Contribution

It introduces an RNN-based approach with an extension to the SoftED metric for better event detection evaluation in force sensor signals.

Findings

BGRU-based RNN outperforms CNNs in paddle stroke detection

Extended SoftED metric effectively assesses event detection in time windows

Automated detection reduces need for human intervention in performance analysis

Abstract

Assessing an athlete's performance in canoe sprint is often established by measuring a variety of kinematic parameters during training sessions. Many of these parameters are related to single or multiple paddle stroke cycles. Determining on- and offset of these cycles in force sensor signals is usually not straightforward and requires human interaction. This paper explores convolutional neural networks (CNNs) and recurrent neural networks (RNNs) in terms of their ability to automatically predict these events. In addition, our work proposes an extension to the recently published SoftED metric for event detection in order to properly assess the model performance on time windows. In our results, an RNN based on bidirectional gated recurrent units (BGRUs) turned out to be the most suitable model for paddle stroke detection.