Abstract:Deep neural network models, especially CNNs, have a wide range of applications in many fields, but their high computational power requirements limit the deployment applications in many resource-constrained embedded devices. Pruning techniques reduce the computational power requirements of models by removing redundant structures from CNNs. Most existing static pruning methods use a global uniform pruning rate to prune pre-trained models and require finetuning to recover accuracy after pruning, resulting in high training costs, and the global uniform pruning rate is sub-optimal. While dynamic pruning methods perform pruning during training and use auxiliary modules to calculate the saliency scores of channels, but do not exploit its function of assisting network training. We propose an adaptive structured continuous sparse network pruning method based on the attention mechanism that prunes the original network during training. The attention-based channel similarity calculation module calculates the channel saliency scores while refining features to assist network training, and the adaptive continuous sparse control module gradually discretizes the channel saliency scores and assigns the pruning rate of each layer according to the preset pruning rate target. The pruned model is output after training and no additional fine-tuning is required. We validate the proposed method on CIFAR-10 and the large-scale dataset Imagenet using networks with different structures, and our method outperforms the comparative pruning methods at different pruning rates. In CIFAR-10, our method can reduce VGG-16 by 34.4% FLOPs while top-1 accuracy increased by 0.19%. In Imagenet, we can reduce ResNet-34 by 51.5% FLOPs, while the top-1 accuracy decreases by only 0.89%.

A Network Pruning Method for Remote Sensing Image Scene Classification

Pruning Convolutional Neural Networks for Remote Sensing Images

Network Pruning for Remote Sensing Images Classification Based on Interpretable CNNs

A lightweight and stochastic depth residual attention network for remote sensing scene classification

Multiobjective ResNet Pruning by Means of EMOAs for Remote Sensing Scene Classification

Remote Sensing Image Target Recognition Based on Pruned Deep Neural Network Models

Energy-Based CNN Pruning for Remote Sensing Scene Classification

Absorption Pruning of Deep Neural Network for Object Detection in Remote Sensing Imagery

A Feature-map Discriminant Perspective for Pruning Deep Neural Networks

Energy-based CNNs Pruning for Remote Sensing Scene Classification

Data Customization-based Multiobjective Optimization Pruning Framework for Remote Sensing Scene Classification

RemoteTrimmer: Adaptive Structural Pruning for Remote Sensing Image Classification

Multi-Objective Net Architecture Pruning for Remote Sensing Classification.

An IQE Criterion-Based Method for SAR Images Classification Network Pruning.

Regularized Training Framework for Combining Pruning and Quantization to Compress Neural Networks

Network Collaborative Pruning Method for Hyperspectral Image Classification Based on Evolutionary Multi-Task Optimization

SNPF: Sensitiveness Based Network Pruning Framework for Efficient Edge Computing

An Efficient and Lightweight Convolutional Neural Network for Remote Sensing Image Scene Classification

Small Network for Lightweight Task in Computer Vision: A Pruning Method Based on Feature Representation

Attention-based Adaptive Structured Continuous Sparse Network Pruning

Network Pruning Using Sparse Learning and Genetic Algorithm