Continual Segmentation under Joint Nonstationarity
It addresses a realistic but underexplored problem in continual learning for dense prediction, providing a benchmark and method for heterogeneous environments.
The paper formalizes continual semantic segmentation under joint nonstationarity (simultaneous class, domain, and label shifts) and proposes gradient-adaptive stabilization with semi-supervised learning to address instability and overfitting. The method consistently outperforms prior approaches across class-incremental, domain-incremental, and few-shot regimes.
Evolving data streams induce joint nonstationarity in continual semantic segmentation, where semantic classes, input distributions, and supervision availability change simultaneously over time. This setting reflects practical structured prediction systems, yet remains largely unexplored in prior continual learning work, which typically studies these factors in isolation. We formalize continual segmentation under coupled class, domain, and label shifts and investigate learning in heterogeneous dense prediction environments with limited annotations and abundant unlabeled data. To address instability and overfitting arising from few-shot supervision under distribution drift, we introduce gradient-adaptive stabilization, a parameter-wise regularization mechanism implemented via gradient-scaled stochastic perturbations that promotes a principled stability-plasticity tradeoff. We further leverage unlabeled data through semi-supervised learning and introduce prototype anchored supervision that validates pseudo-labels via joint confidence and prototype consistency. Together, these mechanisms enable learning under joint nonstationarity in continual segmentation. Extensive empirical evaluation across class-incremental, domain-incremental, and few-shot regimes demonstrates consistent improvements over prior methods in heterogeneous structured prediction settings. Our results expose fundamental failure modes of existing continual segmentation approaches and provide insight into learning robust dense predictors in dynamically evolving environments.