REPrune: Channel Pruning via Kernel Representative Selection

arXiv:2402.17862v1 Announce Type: new
Abstract: Channel pruning is widely accepted to accelerate modern convolutional neural networks (CNNs). The resulting pruned model benefits from its immediate deployment on general-purpose software and hardware resources. However, its large pruning granularity, specifically at the unit of a convolution filter, often leads to undesirable accuracy drops due to the inflexibility of deciding how and where to introduce sparsity to the CNNs. In this paper, we propose REPrune, a novel channel pruning technique that emulates kernel …

abstract accuracy arxiv benefits cnns convolution convolutional neural networks cs.ai cs.cv deployment filter general hardware kernel leads modern networks neural networks pruning resources software type via