DNN-Based Online Source Counting Based on Spatial Generalized Magnitude Squared Coherence

2026-01-28

AI Summary

Accurately estimating the number of active acoustic sources in real time remains a key challenge in acoustic signal processing. This work addresses this problem by formulating it as a frame-level change detection task and proposes a novel approach that integrates spatial coherence analysis with deep learning. Specifically, binaural signals are first preprocessed via spatial whitening, followed by extraction of the spatial generalized magnitude-squared coherence (SGMSC) features. A lightweight neural network then performs online source counting based on these features. Evaluated in simulated reverberant environments with up to four concurrent speakers and background noise, the proposed method demonstrates efficient and robust performance in estimating the number of active sources, making it well-suited for real-time applications in complex acoustic scenarios.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Active Learning
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Web Mining and Content Analysis: Large pretrained models with web data
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The number of active sound sources is a key parameter in many acoustic signal processing tasks, such as source localization, source separation, and multi-microphone speech enhancement. This paper proposes a novel method for online source counting by detecting changes in the number of active sources based on spatial coherence. The proposed method exploits the fact that a single coherent source in spatially white background noise yields high spatial coherence, whereas only noise results in low spatial coherence. By applying a spatial whitening operation, the source counting problem is reformulated as a change detection task, aiming to identify the time frames when the number of active sources changes. The method leverages the generalized magnitude-squared coherence as a measure to quantify spatial coherence, providing features for a compact neural network trained to detect source count changes framewise. Simulation results with binaural hearing aids in reverberant acoustic scenes with up to 4 speakers and background noise demonstrate the effectiveness of the proposed method for online source counting.

Problem

• source counting
• active sound sources
• spatial coherence
• online detection
• acoustic signal processing

Innovation

• source counting
• spatial coherence
• generalized magnitude-squared coherence
• change detection
• deep neural network