用瓦瑟斯坦距离提升未知物体分割准确率
Out-of-Distribution Segmentation via Wasserstein-Based Evidential Uncertainty
- 基于瓦瑟斯坦损失捕捉分布差异，保持概率单纯形结构
- 在多个数据集上显著优于传统不确定性方法
- 适合自动驾驶等需要识别未知障碍物的场景
深度神经网络在语义分割中表现优异，但仅限于预定义类别，在开放世界中遇到未知物体时会失效。在自动驾驶等安全关键应用中，识别并分割分布外（OOD）物体至关重要。本文提出一种基于瓦瑟斯坦损失的证据分割框架，该方法在保留概率单纯形几何结构的同时，有效捕捉分布间距离。结合KL散度正则化与Dice结构一致性项，相比传统不确定性方法，显著提升了对未知物体的分割性能。
原文摘要 · Abstract (English)
Deep neural networks achieve superior performance in semantic segmentation, but are limited to a predefined set of classes, which leads to failures when they encounter unknown objects in open-world scenarios. Recognizing and segmenting these out-of-distribution (OOD) objects is crucial for safety-critical applications such as automated driving. In this work, we present an evidence segmentation framework using a Wasserstein loss, which captures distributional distances while respecting the probability simplex geometry. Combined with Kullback-Leibler regularization and Dice structural consistency terms, our approach leads to improved OOD segmentation performance compared to uncertainty-based approaches.
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。