Analytic Incremental Learning For Sound Source Localization With Imbalance Rectification

2026-01-26

AI Summary

This work addresses the dual imbalance challenge in real-world sound source localization, arising from long-tailed directional distributions within tasks and data shift or overlap across tasks. To tackle this without relying on replay mechanisms, the authors propose a unified incremental learning framework. The approach leverages the peak characteristics of GCC-PHAT for data augmentation to mitigate intra-task imbalance and introduces an analytical dynamic imbalance corrector that enables task-adaptive parameter updates and regularization. The method effectively avoids catastrophic forgetting while achieving state-of-the-art performance on the SSLR benchmark, with 89.0% accuracy, a mean absolute error of 5.3°, and a backward transfer metric of 1.6, significantly outperforming existing approaches.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Learning to Search
• Intelligent Robots: Localization, Mapping, and Navigation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Sound source localization (SSL) demonstrates remarkable results in controlled settings but struggles in real-world deployment due to dual imbalance challenges: intra-task imbalance arising from long-tailed direction-of-arrival (DoA) distributions, and inter-task imbalance induced by cross-task skews and overlaps. These often lead to catastrophic forgetting, significantly degrading the localization accuracy. To mitigate these issues, we propose a unified framework with two key innovations. Specifically, we design a GCC-PHAT-based data augmentation (GDA) method that leverages peak characteristics to alleviate intra-task distribution skews. We also propose an Analytic dynamic imbalance rectifier (ADIR) with task-adaption regularization, which enables analytic updates that adapt to inter-task dynamics. On the SSLR benchmark, our proposal achieves state-of-the-art (SoTA) results of 89.0% accuracy, 5.3{\deg} mean absolute error, and 1.6 backward transfer, demonstrating robustness to evolving imbalances without exemplar storage.

Problem

• sound source localization
• class imbalance
• catastrophic forgetting
• direction-of-arrival
• incremental learning

Innovation

• analytic incremental learning
• imbalance rectification
• sound source localization
• GCC-PHAT-based data augmentation
• task-adaptive regularization