Statistical validation and full-sphere extension of a Bayesian model for human static sound localisation

2026-06-23

AI Summary

This study addresses the lack of principled statistical validation in existing auditory spatial localization models and the limitations of head-related transfer function (HRTF) templates in achieving full-sphere coverage and high-frequency fidelity. The authors propose the first Bayesian sound localization model equipped with an explicit likelihood function, which jointly integrates noise-aware features and individualized HRTFs to infer sound source direction. Model verifiability is achieved through parameter recovery analysis. Systematic evaluation demonstrates that full-sphere HRTF coverage and high-frequency fidelity are more critical than the interpolation algorithm itself. Using behavioral data from 33 participants, the model reliably estimates individual spectral and motion parameters, clarifying the core determinants of HRTF quality and providing both theoretical grounding and open-source tools for auditory localization research.

Abstract

Auditory models are central tools for studying spatial hearing, yet their validation typically relies on heuristic performance metrics rather than principled statistical methods. We present two contributions building on a Bayesian sound localisation model that jointly infers sound direction from noisy perceptual features and individual head-related transfer functions (HRTFs). First, we derive an explicit likelihood function and validate it through parameter recovery on simulated data and fitting to behavioural responses from 33 participants, demonstrating that the framework reliably identifies individual sensorimotor and spectral parameters. Second, we use this framework to compare four HRTF template interpolation methods, showing that full-sphere spatial coverage and high-frequency spectral fidelity are the primary determinants of template quality, while the specific interpolation algorithm is secondary. Together, these results show that standard model-based statistical methods can address both fundamental questions in spatial hearing and applied problems such as perceptual HRTF evaluation. An open-source Python implementation is released alongside this work.

Problem

• sound localisation
• Bayesian model
• HRTF
• statistical validation
• spatial hearing

Innovation

• Bayesian sound localisation
• statistical validation
• HRTF interpolation
• full-sphere coverage
• parameter recovery