DA-Cal: Towards Cross-Domain Calibration in Semantic Segmentation
This addresses a critical issue for safety-critical applications by enhancing calibration in cross-domain semantic segmentation, though it is incremental as it builds on existing self-training frameworks.
The paper tackled the problem of poor calibration in unsupervised domain adaptation for semantic segmentation, where prediction confidence does not align with accuracy, and proposed DA-Cal, a framework that improved target domain calibration and achieved performance gains without inference overhead across multiple benchmarks.
While existing unsupervised domain adaptation (UDA) methods greatly enhance target domain performance in semantic segmentation, they often neglect network calibration quality, resulting in misalignment between prediction confidence and actual accuracy -- a significant risk in safety-critical applications. Our key insight emerges from observing that performance degrades substantially when soft pseudo-labels replace hard pseudo-labels in cross-domain scenarios due to poor calibration, despite the theoretical equivalence of perfectly calibrated soft pseudo-labels to hard pseudo-labels. Based on this finding, we propose DA-Cal, a dedicated cross-domain calibration framework that transforms target domain calibration into soft pseudo-label optimization. DA-Cal introduces a Meta Temperature Network to generate pixel-level calibration parameters and employs bi-level optimization to establish the relationship between soft pseudo-labels and UDA supervision, while utilizing complementary domain-mixing strategies to prevent overfitting and reduce domain discrepancies. Experiments demonstrate that DA-Cal seamlessly integrates with existing self-training frameworks across multiple UDA segmentation benchmarks, significantly improving target domain calibration while delivering performance gains without inference overhead. The code will be released.