Versión en Español ## Entrainment metrics **entrainment-metrics** is a Python library for measuring the degree of *entrainment* in a dialogue, defined as a speaker's tendency to adapt their speech to match their interlocutor's. Given a dialogue between two speakers, this library outputs three entrainment measures (*proximity*, *convergence* and *synchrony*) on a number of acoustic-prosodic features such as pitch, intensity and speech rate. * Repository: https://github.com/erikernst4/entrainment-metrics * Documentation: https://entrainment-metrics.readthedocs.io * Getting Started page: https://entrainment-metrics.readthedocs.io/en/latest/usage/getting_started.html#getting-started * Notebook tutorial with step-by-step guide: https://github.com/erikernst4/entrainment-metrics/blob/master/examples/entrainment-example.ipynb The library was created by Erik Ernst, Ramiro H. Galvez and Agustin Gravano, and is based on two papers: * R.H. Galvez, L. Gauder, J. Luque, A. Gravano (2020). ["A unifying framework for modeling acoustic/prosodic entrainment: definition and evaluation on two large corpora"](https://aclanthology.org/2020.sigdial-1.27/), in Proceedings of SIGDIAL 2020. * J.M. Perez, R.H. Galvez, A. Gravano (2016), ["Disentrainment may be a positive thing: A novel measure of unsigned acoustic-prosodic synchrony, and its relation to speaker engagement"](https://www.isca-speech.org/archive/interspeech_2016/perez16_interspeech.html), in Proceedings of Interspeech 2016. We hope this will be a useful tool for researchers and practitioners working in the fields of speech analysis and spoken dialogue systems. --- **Last updated:** 16 Aug 2023