SNF: Filter Pruning via Searching the Proper Number of Filters

TL;DR

This paper introduces SNF, a filter pruning method that searches for the optimal number of filters per layer, leading to more efficient CNNs with minimal accuracy loss on CIFAR-10 and ImageNet.

Contribution

SNF is the first method to optimize the number of filters per layer during pruning, improving CNN efficiency while maintaining accuracy.

Findings

Achieves state-of-the-art accuracy on CIFAR-10 with significant FLOPs reduction.

Improves Top-1 accuracy on CIFAR-10 with ResNet-110 after pruning.

Maintains competitive accuracy on ImageNet with substantial FLOPs reduction.

Abstract

Convolutional Neural Network (CNN) has an amount of parameter redundancy, filter pruning aims to remove the redundant filters and provides the possibility for the application of CNN on terminal devices. However, previous works pay more attention to designing evaluation criteria of filter importance and then prune less important filters with a fixed pruning rate or a fixed number to reduce convolutional neural networks' redundancy. It does not consider how many filters to reserve for each layer is the most reasonable choice. From this perspective, we propose a new filter pruning method by searching the proper number of filters (SNF). SNF is dedicated to searching for the most reasonable number of reserved filters for each layer and then pruning filters with specific criteria. It can tailor the most suitable network structure at different FLOPs. Filter pruning with our method leads to the state-of-the-art (SOTA) accuracy on CIFAR-10 and achieves competitive performance on ImageNet ILSVRC-2012.SNF based on the ResNet-56 network achieves an increase of 0.14% in Top-1 accuracy at 52.94% FLOPs reduction on CIFAR-10. Pruning ResNet-110 on CIFAR-10 also improves the Top-1 accuracy of 0.03% when reducing 68.68% FLOPs. For ImageNet, we set the pruning rates as 52.10% FLOPs, and the Top-1 accuracy only has a drop of 0.74%. The codes can be available at https://github.com/pk-l/SNF.