PyEyesWeb – movement quality in Python

An open-source Python toolkit that computes features of expressive movement from motion data – from motion capture, video pose estimation or inertial sensors. It continues the EyesWeb platform of Casa Paganini – InfoMus (Antonio Camurri, Gualtiero Volpe) for modern research workflows.

What it computes

Five feature families are implemented; mid- and high-level qualities are under development.

Feature family	What it describes	Example for the violin
Smoothness	how fluid a movement is (spectral arc length SPARC, jerk)	bow changes, string crossings
Synchronisation	phase coupling between body parts or persons (Hilbert phase locking)	bow arm and head; players in an ensemble
Bilateral symmetry	how similar left and right move	functional asymmetry of violin and bow side
Contraction and expansion	how much space the body takes	opening and closing of the arms
Equilibrium	balance of the body over its base	weight shift while playing

How to use it with Healthy Strings data

Install: pip install pyeyesweb (Python; MIT licence).
Load the landmark table (e.g. BlazePose CSV from SInES) with pandas – mind the decimal comma.
Pass time-stamped arrays of the points you need (wrists, elbows, head) to the feature functions; windows of about 0.5–3 s for synchronisation and symmetry.
Collect the feature dictionaries in a table and analyse them per task in R or Python.

PyEyesWeb also runs in real time and sends features via OSC to TouchDesigner, Max/MSP or Unity – for interactive feedback and performance.

Literature

Sabharwal, S. R., Coletta, P., Corbellini, N., Ghisio, S., Romano, G., Al Foysal, A., Volpe, G., & Camurri, A. (2026). PyEyesWeb: An open source toolkit for multimodal movement feature extraction. In Proceedings of the International Conference on Movement and Computing (MOCO ’26), Montpellier. ACM. https://doi.org/10.1145/3802842.3802904
Niewiadomski, R., Kolykhalova, K., Piana, S., Alborno, P., Volpe, G., & Camurri, A. (2019). Analysis of movement quality in full-body physical activities. ACM Transactions on Interactive Intelligent Systems, 9(1), Article 1. https://doi.org/10.1145/3132369
Glowinski, D., Dardard, F., Gnecco, G., Piana, S., & Camurri, A. (2015). Expressive non-verbal interaction in a string quartet: An analysis through head movements. Journal on Multimodal User Interfaces, 9, 55–68. https://doi.org/10.1007/s12193-014-0154-3
Matthiopoulou, O., Gnecco, G., Sanguineti, M., Mottet, D., Bardy, B., & Camurri, A. (2024). Towards the automated analysis of expressive gesture qualities in full-body movement: The perceived origin of movement. Human-centric Computing and Information Sciences, 14. https://doi.org/10.22967/HCIS.2024.14.054

Package: pypi.org/project/pyeyesweb. With Giorgio Gnecco (IMT Lucca) this line of research also analyses musicians’ movements as trajectories over time – head movements and leadership in orchestras and string quartets, and the body region from which a movement originates.