Exercise motion classification from large-scale wearable sensor data using convolutional neural networks

Terry Taewoong Um,Vahid Babakeshizadeh,Dana Kulić,Dana Kulic
DOI: https://doi.org/10.1109/iros.2017.8206051
2017-09-01
Abstract:The ability to accurately identify human activities is essential for developing automatic rehabilitation and sports training systems. In this paper, large-scale exercise motion data obtained from a forearm-worn wearable sensor are classified with a convolutional neural network (CNN). Time-series data consisting of accelerometer and orientation measurements are formatted as images, allowing the CNN to automatically extract discriminative features. A comparative study on the effects of image formatting and different CNN architectures is also presented. The best performing configuration classifies 50 gym exercises with 92.1% accuracy.
What problem does this paper attempt to address?