提升分割模型在域外情况下的可靠性，让不确定度更真实反映错误区域。
Segment Anything with Robust Uncertainty-Accuracy Correlation
- 设计轻量级不确定性头，联合扰动纹理与几何以增强鲁棒性
- 在23个零样本领域中，不确定性与准确率相关性显著提升
- 适合需要高可信分割结果的医疗、自动驾驶场景
尽管SAM具有强大的零样本性能，但在分布外时表现不可靠，原因在于掩码级置信度混淆（MCC），即单一基于IoU的掩码分数无法反映边界附近像素的可靠性。受神经网络对纹理的偏好与人类视觉以形状为中心处理方式的对比启发，我们把域外变化建模为外观变化与非刚性形变的联合影响。提出鲁棒不确定性-准确性关联分割模型（RUAC），在外观与形变漂移下实现像素级不确定性估计。RUAC引入轻量级不确定性头，通过协同风格-形变攻击进行训练，同时采用不确定性-准确性对齐机制，确保在对抗扰动下不确定性仍能准确标出错误像素。在23个零样本领域上，RUAC提升了分割质量，且不确定性-准确性相关性更强。
原文摘要 · Abstract (English)
Despite strong zero-shot performance, SAM is unreliable under domain shift due to Mask-level Confidence Confusion (MCC), where a single IoU-based mask score fails to reflect pixel-wise reliability near boundaries. Motivated by the contrast between texture-biased shortcuts in neural networks and shape-centric processing in human vision, we model out-of-domain variation as appearance shifts and non-rigid deformations that jointly stress calibration. We propose Segment Anything with Robust Uncertainty-Accuracy Correlation (RUAC) for robust pixel-wise uncertainty estimation under appearance and deformation shifts. RUAC adds a lightweight uncertainty head, trains it with a collaborative style-deformation attack that jointly perturbs texture and geometry, and applies Uncertainty-Accuracy Alignment to ensure uncertainty consistently highlights erroneous pixels even under adversarial perturbations. Across 23 zero-shot domains, RUAC improves segmentation quality and yields more faithful uncertainty with stronger uncertainty-accuracy correlation. Project page: https://hongyouzhou.github.io/ruac/.
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。