Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function

Head-Related Transfer Functions (HRTFs) play a vital role in binaural spatial audio rendering. With the release of numerous HRTF datasets in recent years, abundant data has become available to support HRTF-related research based on deep learning. However, measurement discrepancies across different d...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiale Zhao, Dingding Yao, Junfeng Li
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Open Journal of Signal Processing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11082560/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849246292440514560
author Jiale Zhao
Dingding Yao
Junfeng Li
author_facet Jiale Zhao
Dingding Yao
Junfeng Li
author_sort Jiale Zhao
collection DOAJ
description Head-Related Transfer Functions (HRTFs) play a vital role in binaural spatial audio rendering. With the release of numerous HRTF datasets in recent years, abundant data has become available to support HRTF-related research based on deep learning. However, measurement discrepancies across different datasets introduce significant variations in the data and directly merging these datasets may lead to systematic biases. The recent Listener Acoustic Personalization Challenge 2024 (European Signal Processing Conference) dealt with this issue, with the task of harmonizing different datasets to achieve lower classification accuracy while meeting thresholds over various localization metrics. To mitigate cross-dataset differences, this paper proposes a neural network-based HRTF harmonization approach aimed at eliminating dataset-specific properties embedded in the original measurements. The proposed method utilizes a perceptually relevant loss function, which jointly constrains multiple objectives, including interaural level differences, auditory-filter excitation patterns, and classification accuracy. Experimental results based on eight datasets demonstrate that the proposed approach can effectively minimize distributional disparities between datasets while mostly preserving localization performance. The classification accuracy for harmonized HRTFs between different datasets is reduced to as low as 31%, indicating a significant reduction in cross-dataset discrepancies. The proposed method ranked first in this challenge, which validates its effectiveness.
format Article
id doaj-art-e29b1704db104680922ca9c53fbd5355
institution Kabale University
issn 2644-1322
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Open Journal of Signal Processing
spelling doaj-art-e29b1704db104680922ca9c53fbd53552025-08-20T03:58:32ZengIEEEIEEE Open Journal of Signal Processing2644-13222025-01-01686587510.1109/OJSP.2025.359024811082560Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss FunctionJiale Zhao0https://orcid.org/0009-0009-9875-8859Dingding Yao1https://orcid.org/0000-0002-9610-8782Junfeng Li2https://orcid.org/0000-0002-6272-9169Key Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustics, Chinese Academy of Sciences, Beijing, ChinaKey Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustics, Chinese Academy of Sciences, Beijing, ChinaKey Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustics, Chinese Academy of Sciences, Beijing, ChinaHead-Related Transfer Functions (HRTFs) play a vital role in binaural spatial audio rendering. With the release of numerous HRTF datasets in recent years, abundant data has become available to support HRTF-related research based on deep learning. However, measurement discrepancies across different datasets introduce significant variations in the data and directly merging these datasets may lead to systematic biases. The recent Listener Acoustic Personalization Challenge 2024 (European Signal Processing Conference) dealt with this issue, with the task of harmonizing different datasets to achieve lower classification accuracy while meeting thresholds over various localization metrics. To mitigate cross-dataset differences, this paper proposes a neural network-based HRTF harmonization approach aimed at eliminating dataset-specific properties embedded in the original measurements. The proposed method utilizes a perceptually relevant loss function, which jointly constrains multiple objectives, including interaural level differences, auditory-filter excitation patterns, and classification accuracy. Experimental results based on eight datasets demonstrate that the proposed approach can effectively minimize distributional disparities between datasets while mostly preserving localization performance. The classification accuracy for harmonized HRTFs between different datasets is reduced to as low as 31%, indicating a significant reduction in cross-dataset discrepancies. The proposed method ranked first in this challenge, which validates its effectiveness.https://ieeexplore.ieee.org/document/11082560/Head-related transfer functioncross-datasetharmonizationdeep learning
spellingShingle Jiale Zhao
Dingding Yao
Junfeng Li
Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function
IEEE Open Journal of Signal Processing
Head-related transfer function
cross-dataset
harmonization
deep learning
title Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function
title_full Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function
title_fullStr Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function
title_full_unstemmed Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function
title_short Cross-Dataset Head-Related Transfer Function Harmonization Based on Perceptually Relevant Loss Function
title_sort cross dataset head related transfer function harmonization based on perceptually relevant loss function
topic Head-related transfer function
cross-dataset
harmonization
deep learning
url https://ieeexplore.ieee.org/document/11082560/
work_keys_str_mv AT jialezhao crossdatasetheadrelatedtransferfunctionharmonizationbasedonperceptuallyrelevantlossfunction
AT dingdingyao crossdatasetheadrelatedtransferfunctionharmonizationbasedonperceptuallyrelevantlossfunction
AT junfengli crossdatasetheadrelatedtransferfunctionharmonizationbasedonperceptuallyrelevantlossfunction