An auditory model for simulating human sound localisation¶
bayesian_listener is a Python package for simulating and fitting a
Bayesian model of human sound localisation. Given an individual’s
head-related transfer functions (HRTFs) and a binaural sound, it predicts full response
distributions over all source directions (accounting for spectral,
binaural, and motor-noise uncertainties). Listener-specific noise parameters
can be estimated from measured pointing data via maximum-likelihood
optimisation.
Where to start¶
Install the package and run your first localization simulation in under
five minutes. Start here if you are new to bayesian_listener.
Task-oriented walkthroughs: simulate localization responses, fit the model to your own HRTF, and compare interpolation methods.
Complete documentation of every public class, method, and function. Jump here if you know what you need and want parameter details.
The statistical framework, likelihood equations, noise-parameter table, and known limitations. Start here if you are reading the paper.
Citing this work¶
If you use bayesian_listener in your research, please cite the
original model paper and its statistical validation:
@article{barumerli2026,
author = {R. Barumerli and F. Brinkmann and E. Zanoni and A. Hoyer
and L. Picinali and M. Geronazzo},
title = {Statistical validation and full-sphere extension of a {Bayesian}
model for human static sound localisation},
journal = {Submitted to Acta Acustica},
year = {2026},
url = {https://arxiv.org/abs/2606.24367}
}
@article{barumerli2023,
author = {Barumerli, Roberto and Majdak, Piotr and Geronazzo, Michele
and Meijer, Demi and Avanzini, Federico and Baumgartner, Robert},
title = {A {Bayesian} model for human directional localization of
broadband static sound sources},
journal = {Acta Acustica},
volume = {7},
pages = {12},
year = {2023},
doi = {10.1051/aacus/2023006},
}