Head-Related Transfer Function Interpolation from Spatially Sparse Measurements Using Autoencoder with Source Position Conditioning

TL;DR

This paper introduces an autoencoder-based approach for interpolating head-related transfer functions (HRTFs) from sparse measurements, leveraging source position conditioning to improve accuracy and generalization to unseen subjects.

Contribution

The paper presents a novel autoencoder architecture with source position conditioning and an aggregation module, enabling effective HRTF interpolation from limited measurements and generalization across subjects.

Findings

Achieves comparable interpolation performance with only one-eighth of measurements.

Works well for unseen subjects, demonstrating good generalization.

Outperforms traditional RLR-based methods in sparse measurement scenarios.

Abstract

We propose a method of head-related transfer function (HRTF) interpolation from sparsely measured HRTFs using an autoencoder with source position conditioning. The proposed method is drawn from an analogy between an HRTF interpolation method based on regularized linear regression (RLR) and an autoencoder. Through this analogy, we found the key feature of the RLR-based method that HRTFs are decomposed into source-position-dependent and source-position-independent factors. On the basis of this finding, we design the encoder and decoder so that their weights and biases are generated from source positions. Furthermore, we introduce an aggregation module that reduces the dependence of latent variables on source position for obtaining a source-position-independent representation of each subject. Numerical experiments show that the proposed method can work well for unseen subjects and achieve an interpolation performance with only one-eighth measurements comparable to that of the RLR-based method.