在CNNs中促进形状偏置：用于损坏鲁棒性的基于频率和对比的正则化
Promoting Shape Bias in CNNs: Frequency-Based and Contrastive Regularization for Corruption Robustness
AI总结：
针对CNNs依赖纹理导致对常见损坏脆弱的问题，提出基于频率滤波和监督对比学习的两种正则化策略，在CIFAR-10-C上提升了损坏鲁棒性且不损失干净准确率。
摘要：
卷积神经网络（CNNs）在图像分类方面表现出色，但仍然容易受到人类能轻松应对的常见损坏的影响。导致这种脆弱性的一个关键原因是它们依赖局部纹理线索而非全局物体形状——这与人类感知形成鲜明对比。为了解决这个问题，我们提出了两种互补的正则化策略，旨在鼓励形状偏置表示并增强鲁棒性。第一种引入辅助损失，强制原始输入与低频滤波输入之间的特征一致性，抑制对高频纹理的依赖。第二种结合监督对比学习，围绕类一致、形状相关的表示来构建特征空间。在CIFAR-10-C基准上评估，两种方法均提升了损坏鲁棒性，且未降低干净准确率。我们的结果表明，损失级正则化能有效引导CNNs形成更具形状感知、更具弹性的表示。
英文摘要：
Convolutional Neural Networks (CNNs) excel at image classification but remain vulnerable to common corruptions that humans handle with ease. A key reason for this fragility is their reliance on local texture cues rather than global object shapes -- a stark contrast to human perception. To address this, we propose two complementary regularization strategies designed to encourage shape-biased representations and enhance robustness. The first introduces an auxiliary loss that enforces feature consistency between original and low-frequency filtered inputs, discouraging dependence on high-frequency textures. The second incorporates supervised contrastive learning to structure the feature space around class-consistent, shape-relevant representations. Evaluated on the CIFAR-10-C benchmark, both methods improve corruption robustness without degrading clean accuracy. Our results suggest that loss-level regularization can effectively steer CNNs toward more shape-aware, resilient representations.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
