Learning Instance-Specific Adaptation for Cross-Domain Segmentation

TL;DR

This paper introduces a simple test-time adaptation method for cross-domain image segmentation that calibrates BatchNorm statistics on individual instances without needing target domain data during training, improving performance significantly.

Contribution

It proposes a learnable, instance-specific BatchNorm calibration method combined with data augmentation, eliminating the need for target domain data or costly test-time training.

Findings

Achieves significant performance improvements over existing methods.

Reaches new state-of-the-art results when combined with domain generalization techniques.

Does not require access to target domain data during training or expensive optimization.

Abstract

We propose a test-time adaptation method for cross-domain image segmentation. Our method is simple: Given a new unseen instance at test time, we adapt a pre-trained model by conducting instance-specific BatchNorm (statistics) calibration. Our approach has two core components. First, we replace the manually designed BatchNorm calibration rule with a learnable module. Second, we leverage strong data augmentation to simulate random domain shifts for learning the calibration rule. In contrast to existing domain adaptation methods, our method does not require accessing the target domain data at training time or conducting computationally expensive test-time model training/optimization. Equipping our method with models trained by standard recipes achieves significant improvement, comparing favorably with several state-of-the-art domain generalization and one-shot unsupervised domain adaptation approaches. Combining our method with the domain generalization methods further improves performance, reaching a new state of the art.