Temporal-Spectral Alignment with Frequency Adaptation for Source-Free Time-Series Adaptation
2026-06-22 • Machine Learning
Machine Learning
AI summaryⓘ
The authors address a problem where a model trained on one set of time-series data (source) needs to work well on new, unlabeled data (target) without accessing the original source data. They point out that previous methods often miss differences in the frequency content of the data. To fix this, the authors created a method called SAFA that looks at both time patterns and frequency features of the source data. Their method adjusts the frequency of the target signals to better match the source, improving adaptation. They tested SAFA on several datasets and found it works well.
source-free domain adaptationtime-series datafeature shifttemporal driftspectral shiftfrequency domainfrequency adaptationtemporal dependenciesphase and amplitude modulation
Authors
Shichang Meng, Linquan Wu, Xuan Ai, Linqi Song
Abstract
The goal of source-free domain adaptation (SFDA) for time-series data is to transfer knowledge from a pre-trained source model to an unlabeled target domain without requiring access to source data, while addressing feature shift and temporal drift inherent in the signals. Although existing approaches have explored temporal dynamics in unsupervised source-free adaptation, they largely overlook spectral shifts in time-series data. Towards this end, we propose a novel approach termed temporal-Spectral Alignment with Frequency Adaptation (SAFA) for source-free time-series domain adaptation. Specifically, we first model the source domain at multiple scales by jointly capturing temporal dependencies and spectral characteristics. To adapt time-series data in the target domain, we introduce a trainable frequency adaptation module that modulates the phase and amplitude of target signals in the frequency domain to align them with the source distribution. Extensive experiments on multiple benchmark datasets demonstrate the efficacy and robustness of SAFA.