TransAdapt: A Transformative Framework for Online Test Time Adaptive Semantic Segmentation
This work addresses the challenge of adapting semantic segmentation models to unlabeled target domain images in an online setting, which is incremental as it builds on existing test-time adaptation methods.
The paper tackles the problem of online test-time adaptive semantic segmentation by proposing TransAdapt, a framework that uses transformer and input transformations to improve segmentation performance without requiring test-time online training, resulting in up to 17.6% and 2.8% mIOU improvement over no-adaptation and competitive baselines.
Test-time adaptive (TTA) semantic segmentation adapts a source pre-trained image semantic segmentation model to unlabeled batches of target domain test images, different from real-world, where samples arrive one-by-one in an online fashion. To tackle online settings, we propose TransAdapt, a framework that uses transformer and input transformations to improve segmentation performance. Specifically, we pre-train a transformer-based module on a segmentation network that transforms unsupervised segmentation output to a more reliable supervised output, without requiring test-time online training. To also facilitate test-time adaptation, we propose an unsupervised loss based on the transformed input that enforces the model to be invariant and equivariant to photometric and geometric perturbations, respectively. Overall, our framework produces higher quality segmentation masks with up to 17.6% and 2.8% mIOU improvement over no-adaptation and competitive baselines, respectively.