Efficient Unsupervised Domain Adaptation Regression for Spatial-Temporal Sensor Fusion

TL;DR

This paper introduces a scalable unsupervised domain adaptation method for regression in sensor fusion, improving robustness and transferability across environmental and biomedical sensor data without needing labeled target data.

Contribution

The paper presents a novel UDA approach integrated with Spatial-Temporal Graph Neural Networks that aligns inverse Gram matrices, enabling effective domain adaptation for sensor regression tasks.

Findings

Achieves state-of-the-art results on air quality and EEG datasets.

Effectively handles sensor drift and distribution shifts.

No labeled target data required for adaptation.

Abstract

The growing deployment of low-cost, distributed sensor networks in environmental and biomedical domains has enabled continuous, large-scale health monitoring. However, these systems often face challenges related to degraded data quality caused by sensor drift, noise, and insufficient calibration -- factors that limit their reliability in real-world applications. Traditional machine learning methods for sensor fusion and calibration rely on extensive feature engineering and struggle to capture spatial-temporal dependencies or adapt to distribution shifts across varying deployment conditions. To address these challenges, we propose a novel unsupervised domain adaptation (UDA) method tailored for regression tasks. Our proposed method integrates effectively with Spatial-Temporal Graph Neural Networks and leverages the alignment of perturbed inverse Gram matrices between source and target domains, drawing inspiration from Tikhonov regularization. This approach enables scalable and efficient domain adaptation without requiring labeled data in the target domain. We validate our novel method on real-world datasets from two distinct applications: air quality monitoring and EEG signal reconstruction. Our method achieves state-of-the-art performance which paves the way for more robust and transferable sensor fusion models in both environmental and physiological contexts. Our code is available at https://github.com/EPFL-IMOS/TikUDA.