RANP: Resource Aware Neuron Pruning at Initialization for 3D CNNs

TL;DR

This paper introduces RANP, a novel method for pruning 3D CNNs at initialization based on neuron importance and resource consumption, significantly reducing computational costs with minimal accuracy loss.

Contribution

The paper proposes a resource-aware neuron pruning algorithm for 3D CNNs at initialization, enabling high sparsity levels without iterative optimization.

Findings

Achieves 50-95% FLOPs reduction in 3D semantic segmentation tasks.

Reduces memory usage by 35-80% with negligible accuracy loss.

Pruned networks are scalable and transferable to new datasets.

Abstract

Although 3D Convolutional Neural Networks (CNNs) are essential for most learning based applications involving dense 3D data, their applicability is limited due to excessive memory and computational requirements. Compressing such networks by pruning therefore becomes highly desirable. However, pruning 3D CNNs is largely unexplored possibly because of the complex nature of typical pruning algorithms that embeds pruning into an iterative optimization paradigm. In this work, we introduce a Resource Aware Neuron Pruning (RANP) algorithm that prunes 3D CNNs at initialization to high sparsity levels. Specifically, the core idea is to obtain an importance score for each neuron based on their sensitivity to the loss function. This neuron importance is then reweighted according to the neuron resource consumption related to FLOPs or memory. We demonstrate the effectiveness of our pruning method on 3D semantic segmentation with widely used 3D-UNets on ShapeNet and BraTS'18 as well as on video classification with MobileNetV2 and I3D on UCF101 dataset. In these experiments, our RANP leads to roughly 50-95 reduction in FLOPs and 35-80 reduction in memory with negligible loss in accuracy compared to the unpruned networks. This significantly reduces the computational resources required to train 3D CNNs. The pruned network obtained by our algorithm can also be easily scaled up and transferred to another dataset for training.