Weakly-supervised continual learning for class-incremental segmentation
This addresses incremental learning for remote sensing segmentation, but appears incremental as it compares existing regularization terms and pseudo-label strategies.
The paper tackles the problem of adapting semantic segmentation models to new classes in remote sensing using weak supervision, and shows experimental relevance on three public datasets with open-source code.
Transfer learning is a powerful way to adapt existing deep learning models to new emerging use-cases in remote sensing. Starting from a neural network already trained for semantic segmentation, we propose to modify its label space to swiftly adapt it to new classes under weak supervision. To alleviate the background shift and the catastrophic forgetting problems inherent to this form of continual learning, we compare different regularization terms and leverage a pseudo-label strategy. We experimentally show the relevance of our approach on three public remote sensing datasets. Code is open-source and released in this repository: https://github.com/alteia-ai/ICSS}{https://github.com/alteia-ai/ICSS.