Kernel Normalized Convolutional Networks

arXiv:2205.10089v4 Announce Type: replace
Abstract: Existing convolutional neural network architectures frequently rely upon batch normalization (BatchNorm) to effectively train the model. BatchNorm, however, performs poorly with small batch sizes, and is inapplicable to differential privacy. To address these limitations, we propose the kernel normalization (KernelNorm) and kernel normalized convolutional layers, and incorporate them into kernel normalized convolutional networks (KNConvNets) as the main building blocks. We implement KNConvNets corresponding to the state-of-the-art ResNets while forgoing the BatchNorm layers. Through extensive experiments, …

abstract architectures arxiv convolutional neural network cs.cv cs.lg differential differential privacy kernel limitations network networks neural network normalization privacy small them train type