SHAMaNS: Sound Localization with Hybrid Alpha-Stable Spatial Measure and Neural Steerer

2025-06-23

AI Summary

To address the insufficient robustness of multi-source localization under non-Gaussian noise—specifically α-stable noise (α ∈ (0,2))—this paper proposes a novel physics-informed deep learning framework. First, we design a physically grounded neural steered network (“Neural Steerer”) that enables continuous, differentiable interpolation of steering vectors. Second, we introduce an α-stable spatial measure to explicitly model signal spatial characteristics, leveraging its heavy-tailed property to absorb neural interpolation residuals and thereby enhance robustness against impulsive noise. By abandoning the restrictive Gaussian assumption, our approach achieves significantly improved DOA estimation accuracy in overlapping multi-source scenarios. Quantitative evaluation demonstrates consistent superiority over state-of-the-art methods across standard metrics—including DOA error and F1-score—validating both theoretical soundness and practical effectiveness.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

This paper describes a sound source localization (SSL) technique that combines an $α$-stable model for the observed signal with a neural network-based approach for modeling steering vectors. Specifically, a physics-informed neural network, referred to as Neural Steerer, is used to interpolate measured steering vectors (SVs) on a fixed microphone array. This allows for a more robust estimation of the so-called $α$-stable spatial measure, which represents the most plausible direction of arrival (DOA) of a target signal. As an $α$-stable model for the non-Gaussian case ($α$ $in$ (0, 2)) theoretically defines a unique spatial measure, we choose to leverage it to account for residual reconstruction error of the Neural Steerer in the downstream tasks. The objective scores indicate that our proposed technique outperforms state-of-the-art methods in the case of multiple sound sources.

Problem

• Robust sound source localization using hybrid techniques
• Neural network interpolation for steering vectors
• Improved direction estimation in non-Gaussian conditions

Innovation

• Combines α-stable model with neural network
• Uses Neural Steerer for steering vector interpolation
• Leverages α-stable measure for robust DOA estimation