Cross-domain Few-shot Segmentation with Transductive Fine-tuning
This addresses domain adaptation for few-shot segmentation, enabling models to generalize to unseen domains without additional labels, which is incremental but practical for applications like medical imaging.
The paper tackles the problem of few-shot segmentation failing due to domain gaps between base and novel classes by proposing transductive fine-tuning that aligns prototypes with an uncertainty-aware contrastive loss, resulting in consistent and significant performance improvements across natural, remote sensing, and medical image tasks.
Few-shot segmentation (FSS) expects models trained on base classes to work on novel classes with the help of a few support images. However, when there exists a domain gap between the base and novel classes, the state-of-the-art FSS methods may even fail to segment simple objects. To improve their performance on unseen domains, we propose to transductively fine-tune the base model on a set of query images under the few-shot setting, where the core idea is to implicitly guide the segmentation of query images using support labels. Although different images are not directly comparable, their class-wise prototypes are desired to be aligned in the feature space. By aligning query and support prototypes with an uncertainty-aware contrastive loss, and using a supervised cross-entropy loss and an unsupervised boundary loss as regularizations, our method could generalize the base model to the target domain without additional labels. We conduct extensive experiments under various cross-domain settings of natural, remote sensing, and medical images. The results show that our method could consistently and significantly improve the performance of prototypical FSS models in all cross-domain tasks.