让CNN更关注物体形状而非纹理，提升抗图像退化能力
Promoting Shape Bias in CNNs: Frequency-Based and Contrastive Regularization for Corruption Robustness
- 用低频滤波输入与原图的特征一致性约束，抑制对纹理的依赖
- 引入有监督对比学习，构建以形状为中心的特征空间结构
- 在CIFAR-10-C上显著提升抗退化性能，且不牺牲干净图像准确率
卷积神经网络（CNN）在图像分类上表现优异，但对人类轻易应对的常见图像退化仍很脆弱。主要原因是其过度依赖局部纹理线索，而非全局物体形状——这与人类感知形成鲜明对比。为此，我们提出两种互补的正则化策略，旨在促进形状偏向的表征并增强鲁棒性。第一种方法引入辅助损失，强制原始输入与低频滤波输入之间的特征一致性，从而削弱对高频纹理的依赖。第二种方法结合有监督对比学习，将特征空间围绕类别一致、形状相关的表示进行组织。在CIFAR-10-C基准上的评估表明，两种方法均在不降低干净图像准确率的前提下提升了抗退化能力。结果表明，损失级正则化可有效引导CNN向更具形状感知力和鲁棒性的表征演进。
原文摘要 · Abstract (English)
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.
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。