Abstract
Wearing inappropriate running shoes may lead to unnecessary injury through continued strain upon the lower extremities; potentially damaging a runner's performance. Many technologies have been developed for accurate shoe recommendation, which centre on running gait analysis. However, these often require supervised use in the laboratory/shop or exhibit too high a cost for personal use. This work addresses the need for a deployable, inexpensive product with the ability to accurately assess running shoe-type recommendation. This was achieved through quantitative analysis of the running gait from 203 individuals through use of a tri-axial accelerometer and tri-axial gyroscope-based wearable (Mymo). In combination with a custom neural network to provide the shoe-type classifications running within the cloud, we experience an accuracy of 94.6% in classifying the correct type of shoe across unseen test data.
Highlights
Running is one of the most common forms of exercise due to its ease of access, low cost and beneficial health effects [1], [2]
The following section will describe the final results obtained from each facet of the shoe recommendation system in detail; with the ensemble model’s summation and optimisation strategy discussed
GAIT FEATURES: PRONATION AND FOOT STRIKE Foot pronation and foot-strike location algorithms were tested on all 203 datasets, i.e. the algorithm is static and does not benefit from training data
Summary
Running is one of the most common forms of exercise due to its ease of access, low cost and beneficial health effects [1], [2]. Popularity in novice and recreational running has been recently fuelled by the global phenomenon of mass group events [5]. With a growing number of novice and recreational participant’s, rates of running injuries increases with relatively long periods (up to 52 weeks) of injury sustained [6]. This has an obvious economic impact on healthcare utilisation. With advances in computer vision and pattern recognition methods, gait assessment can be automated [21] Such approaches are computationally intensive and cannot be used at scale given the e.g. requirement for multiple cameras
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
