Drainage：一个应对类别不确定性的统一框架
Drainage: A Unifying Framework for Addressing Class Uncertainty
在分类网络输出端加一个可微的 drainage 节点，把概率质量重新分配给不确定性，为噪声标签、实例相关/非对称噪声和 OOD 样本提供统一的拒识出口；做噪声标签或开集识别时可直接替换 softmax 头，并顺带获得数据集清洗能力。
原文依据 we propose a unified framework based on the concept of a "drainage node" which we add at the output of the network
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论现代深度学习在噪声标签、类别歧义以及稳健拒识分布外或损坏样本等方面面临重大挑战。在本工作中，我们提出了一个基于「drainage 节点」概念的统一框架，该节点被添加在网络的输出端。该节点用于将概率质量重新分配给不确定性，同时保留端到端训练与可微性等理想性质。这一机制为高度歧义、异常或带噪的样本提供了自然的出口，尤其适用于实例相关噪声与非对称标签噪声。在向 CIFAR-10/100 标签中加入不同比例的实例相关噪声或非对称噪声的系统性实验中，我们的 drainage 公式在高噪声情形下相比现有方法取得了最高 10% 的准确率提升。在 mini-WebVision、mini-ImageNet 和 Clothing-1M 等真实数据集上，我们的结果达到或超过了现有最优方法。定性分析揭示出一种去噪效应：drainage 神经元持续吸收损坏、错误标注或离群的数据，从而带来更稳定的决策边界。此外，我们的 drainage 公式使应用远超分类任务，对网络规模、半监督数据集清洗以及开集应用都有直接的助益。
英文原文摘要 1394 字符
Modern deep learning faces significant challenges with noisy labels, class ambiguity, as well as the need to robustly reject out-of-distribution or corrupted samples. In this work, we propose a unified framework based on the concept of a "drainage node" which we add at the output of the network. The node serves to reallocate probability mass toward uncertainty, while preserving desirable properties such as end-to-end training and differentiability. This mechanism provides a natural escape route for highly ambiguous, anomalous, or noisy samples, particularly relevant for instance-dependent and asymmetric label noise. In systematic experiments involving the addition of varying proportions of instance-dependent noise or asymmetric noise to CIFAR-10/100 labels, our drainage formulation achieves an accuracy increase of up to 10% over existing approaches in the high-noise regime. Our results on real-world datasets, such as mini-WebVision, mini-ImageNet and Clothing-1M, match or surpass existing state-of-the-art methods. Qualitative analysis reveals a denoising effect, where the drainage neuron consistently absorbs corrupt, mislabeled, or outlier data, leading to more stable decision boundaries. Furthermore, our drainage formulation enables applications well beyond classification, with immediate benefits for web-scale, semi-supervised dataset cleaning, and open-set applications.