GMM-COMET: Continual Source-Free Universal Domain Adaptation via a Mean Teacher and Gaussian Mixture Model-Based Pseudo-Labeling
This addresses the challenge of domain shifts in real-world applications where data evolves over time, but it is incremental as it builds upon existing methods for online source-free universal domain adaptation.
The paper tackles the problem of adapting neural networks to multiple unlabeled target domains sequentially without access to source data, known as continual source-free universal domain adaptation, and introduces GMM-COMET, which combines Gaussian mixture model-based pseudo-labeling with a mean teacher framework to achieve consistent improvements over the source-only model across all evaluated scenarios.
Unsupervised domain adaptation tackles the problem that domain shifts between training and test data impair the performance of neural networks in many real-world applications. Thereby, in realistic scenarios, the source data may no longer be available during adaptation, and the label space of the target domain may differ from the source label space. This setting, known as source-free universal domain adaptation (SF-UniDA), has recently gained attention, but all existing approaches only assume a single domain shift from source to target. In this work, we present the first study on continual SF-UniDA, where the model must adapt sequentially to a stream of multiple different unlabeled target domains. Building upon our previous methods for online SF-UniDA, we combine their key ideas by integrating Gaussian mixture model-based pseudo-labeling within a mean teacher framework for improved stability over long adaptation sequences. Additionally, we introduce consistency losses for further robustness. The resulting method GMM-COMET provides a strong first baseline for continual SF-UniDA and is the only approach in our experiments to consistently improve upon the source-only model across all evaluated scenarios. Our code is available at https://github.com/pascalschlachter/GMM-COMET.