Unified Class and Domain Incremental Learning with Mixture of Experts for Indoor Localization

2025-11-21

AI Summary

In indoor localization, device heterogeneity induces domain shift, while environmental evolution causes class shift, rendering static models obsolete over time. To address these challenges, we propose the first lightweight, sustainable localization framework that jointly handles class-incremental and domain-incremental learning. Methodologically, it integrates a Mixture-of-Experts (MoE) architecture with an ETF-based gating mechanism to enable low-latency, high-accuracy dynamic expert routing and adaptive inference; further, it synergizes incremental learning with domain adaptation to support continuous deployment across heterogeneous devices and evolving environments. Experiments demonstrate that our framework reduces average localization error by 25.6×, worst-case error by 44.5×, and catastrophic forgetting by 21.5× compared to state-of-the-art approaches—marking significant advances in robustness, efficiency, and long-term adaptability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Intelligent Robots: Localization, Mapping, and Navigation
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Indoor localization using machine learning has gained traction due to the growing demand for location-based services. However, its long-term reliability is hindered by hardware/software variations across mobile devices, which shift the model's input distribution to create domain shifts. Further, evolving indoor environments can introduce new locations over time, expanding the output space to create class shifts, making static machine learning models ineffective over time. To address these challenges, we propose a novel unified continual learning framework for indoor localization called MOELO that, for the first time, jointly addresses domain-incremental and class-incremental learning scenarios. MOELO enables a lightweight, robust, and adaptive localization solution that can be deployed on resource-limited mobile devices and is capable of continual learning in dynamic, heterogeneous real-world settings. This is made possible by a mixture-of-experts architecture, where experts are incrementally trained per region and selected through an equiangular tight frame based gating mechanism ensuring efficient routing, and low-latency inference, all within a compact model footprint. Experimental evaluations show that MOELO achieves improvements of up to 25.6x in mean localization error, 44.5x in worst-case localization error, and 21.5x lesser forgetting compared to state-of-the-art frameworks across diverse buildings, mobile devices, and learning scenarios.

Problem

• Addresses domain shifts from mobile device hardware variations
• Solves class shifts from new locations in evolving environments
• Enables continual learning for indoor localization on resource-limited devices

Innovation

• Mixture-of-experts architecture enables incremental learning
• Equiangular tight frame gating ensures efficient expert routing
• Compact model supports low-latency inference on mobile devices