The Microsoft Kinect depth sensor could offer a convenient, markerless solution for quantifying the head and torso movements of pianists to examine the impact of somatic training on playing postures and movement. To assess the suitability of the Kinect for this application, we tracked four professional piano teachers performing scales immediately before and after a week-long workshop involving daily Feldenkrais Awareness through Movement (ATM) lessons. We compared Kinect skeletal tracking data with 2D reference data obtained simultaneously using Dartfish video analysis software. Analysis revealed frequent tracking errors in the Kinect data compared to reference data from Dartfish. Differences in pre- and post-test measurements of forward head position, head height, C7 vertebra height and shoulder displacement did not correspond between Dartfish and Kinect. Our results suggest that one Kinect sensor does not provide enough accuracy to track torso movements of pianists for the purposes of ergonomic assessment in response to somatic training.

Additional Metadata
Keywords Dartfish, Feldenkrais, Kinect, Motion tracking, Piano pedagogy, Posture
Persistent URL dx.doi.org/10.1386/jmte.10.1.51_1
Journal Journal of Music, Technology and Education
Citation
Beacon, J.F. (Jillian F.), Comeau, G. (Gilles), Payeur, P. (Pierre), & Russell, D. (2017). Assessing the suitability of Kinect for measuring the impact of a weeklong Feldenkrais method workshop on pianists’ posture and movement. Journal of Music, Technology and Education, 10(1), 51–72. doi:10.1386/jmte.10.1.51_1