Spatial Self-Attention Driven Feature Re-Weighting for Sound Source Localization in Challenging Acoustic Environments
IEEE Access · 2026 · 14 · 164871-164886
DOI: 10.1109/access.2026.3746902 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Ye Tian und 2 weiteren Autor:innen. Die Arbeit nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Major Science and Technology Support Program of Hebei Province, Open Foundation of the Key Laboratory of Life Search and Rescue Technology for Earthquake and Geological Disaster, Ministry of Emergency Management of China, Graduate Student Science and Technology Innovation Basic Project.
Abstract (Originalsprache)
Sound source localization remains vulnerable in challenging acoustic environments characterized by reverberation and low signal-to-noise ratios. To address this issue, this paper presents spatial self-attention-driven feature re-weighting within a convolutional neural network (SSA-CNN). The re-weighting mechanism combines channel-wise average pooling and channel-wise max pooling to derive complementary spatial statistics and then generates an adaptive weight mask for acoustic feature maps. A training dataset is constructed from the LibriSpeech corpus, and comparative experiments are conducted against GCC-PHAT, SRP-PHAT, TFnet, SELDnet, and MSAT under controlled acoustic conditions and LOCATA-based evaluation. Experimental results show that SSA-CNN maintains a root mean squared angular error (RMSAE) of 7°–15° across multiple reverberation-time (T60) and signal-to-noise-ratio (SNR) settings in the controlled tests. Under the challenging condition T $60=1.5$ s and SNR = 5 dB, SSA-CNN achieves an error of 23.09°, which is lower than the errors obtained by the evaluated baselines under the adopted protocol. These results demonstrate the effectiveness of spatial self-attention-driven feature re-weighting in improving the accuracy and interference robustness of three-dimensional sound source localization.
Förderung
- Major Science and Technology Support Program of Hebei Province
- Open Foundation of the Key Laboratory of Life Search and Rescue Technology for Earthquake and Geological Disaster, Ministry of Emergency Management of China
- Graduate Student Science and Technology Innovation Basic Project
- Basic Scientific Research Business Expense of Central Universities
Zitieren
APA
Tian, Y., Tian, L., & Feng, Y. (2026). Spatial Self-Attention Driven Feature Re-Weighting for Sound Source Localization in Challenging Acoustic Environments. IEEE Access, 14, 164871-164886. https://doi.org/10.1109/access.2026.3746902
Vancouver
Tian Y, Tian L, Feng Y. Spatial Self-Attention Driven Feature Re-Weighting for Sound Source Localization in Challenging Acoustic Environments. IEEE Access. 2026;14:164871-164886. doi:10.1109/access.2026.3746902
BibTeX
@article{Tian2026,
author = {Tian, Ye and Tian, Liqin and Feng, Yanru},
title = {Spatial Self-Attention Driven Feature Re-Weighting for Sound Source Localization in Challenging Acoustic Environments},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {164871--164886},
doi = {10.1109/access.2026.3746902}
}
EUVIMED