通过不确定性建模提升跨域少样本分割的原型判别可靠性
DAUPNet: Domain-Aware Uncertainty Modeling for Reliable Prototype Discrimination in Cross-Domain Few-Shot Semantic Segmentation
- 用分层特征对齐增强支持与查询特征的稳定性
- 概率化表示前景与背景原型，1-shot和5-shot下mIoU分别达72.6%和76.7%
- 将不确定度融入对比优化，适合医学等高域偏移场景
跨域少样本语义分割（CD-FSS）通常依赖于学习域不变表征或改善支持-查询对应关系。然而，严重的域偏移仍使原型匹配不可靠：层级响应不一致会破坏支持表征，确定性原型无法表达边界与外观模糊性，且在优化中对不同可靠性的原型同等对待削弱了前景-背景分离。为此，我们提出DAUPNet，将跨域原型匹配重新定义为不确定性感知的原型判别。DAUPNet首先对齐分层支持-查询特征以提供稳定证据，然后概率化表示前景与背景原型，并利用其估计的不确定性调节对比优化。在四个标准目标域上，DAUPNet在1-shot和5-shot设置下的平均mIoU分别达到72.6%和76.7%，在两个医学数据集上均有显著提升。结果表明，在严重域偏移下，建模原型不确定性并将其融入优化，是一种鲁棒且可解释的CD-FSS方法。代码已开源。
原文摘要 · Abstract (English)
Cross-domain few-shot semantic segmentation (CD-FSS) has predominantly been formulated as learning domain-invariant representations or improving support-query correspondence. Nevertheless, large domain shifts still make prototype matching unreliable: inconsistent hierarchical responses corrupt the support representation, deterministic prototypes cannot express boundary and appearance ambiguity, and treating prototypes with different reliability equally during optimization weakens foreground-background separation. We therefore propose DAUPNet, a unified framework that reformulates cross-domain prototype matching as uncertainty-aware prototype discrimination. DAUPNet first harmonizes hierarchical support-query features to provide stable evidence, then represents foreground and background prototypes probabilistically, and finally uses their estimated uncertainty to regulate contrastive optimization. On four standard target domains, DAUPNet achieves 72.6% and 76.7% average mIoU in the 1-shot and 5-shot settings, respectively, including substantial gains on the two medical domains. These results demonstrate that modeling prototype uncertainty and incorporating it into optimization provides a robust and interpretable approach to CD-FSS under severe domain shift. The code is available at https://github.com/madness-Lei/DAUPNet
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。