Convolutional layers in CNNs implement linear filters which decompose the input into different frequency bands. However, most modern architectures neglect standard principles of filter design when optimizing their model choices regarding the size and shape of the convolutional kernel. In this work, we consider the well-known problem of spectral leakage caused by windowing artifacts in filtering operations in the context of CNNs. We show that the small size of CNN kernels make them susceptible to spectral leakage, which may induce performance-degrading artifacts. To address this issue, we propose the use of larger kernel sizes along with the Hamming window function to alleviate leakage in CNN architectures. We demonstrate improved classification accuracy on multiple benchmark datasets including Fashion-MNIST, CIFAR-10, CIFAR-100 and ImageNet with the simple use of a standard window function in convolutional layers. Finally, we show that CNNs employing the Hamming window display increased robustness against various adversarial attacks.
翻译:CNN 的连锁层采用线性过滤器,分解输入到不同频带中的输入。 然而,大多数现代建筑在优化关于进动内核大小和形状的模型选择时,忽视了过滤设计的标准原则。 在这项工作中,我们考虑到在CNN 的过滤操作中,由窗口艺术品造成的众所周知的光谱渗漏问题。我们显示CNN 的内核体小使其易受光性泄漏的影响,这可能导致性能降解的文物。为了解决这一问题,我们提议使用更大的内核尺寸以及哈明窗口功能来减少CNN 结构中的泄漏。我们展示了多种基准数据集的分类准确性,包括时装-MNIST、CIFAR-10、CIFAR-100和图像网络,在进动层中简单使用标准窗口功能。最后,我们显示,使用哈明窗口的CNN在各种对抗性攻击中表现出了更强的强度。