Multi-Domain Clustering Pruning: Exploring Space and Frequency Similarity Based on GAN (MDCP)

🌟 Features

Static pruning considering both Sapce and Frequency redundancy.

• Network compression plays an important role in accelerating deep neural networks, especially in the application of edge devices such as unmanned cars and drones. Recently, pruning-based methods have been improved significantly, but they still suffer from low efficiency because most of them only pay attention to feature similarities in the space domain. In this paper, we propose a multi-domain structured pruning method based on clustering (MDCP) which seamlessly integrates sufficient information extraction and knowledge distillation within a GAN-based framework, to address these aforementioned limitations.

📌 To do

• More pre-trained models
• running instructions

Several pre-trained models on CIFAR-10 can be download here: ResNet20, ResNet34, ResNet56, ResNet110.

Checkpoints for pruning ResNet on ImageNet released here~

Acknowledgments

• FPGM
• GAL

Citation

If you find this work useful for your research, please cite us:

@article{zhang2023multi,
  title={Multi-Domain Clustering Pruning: Exploring Space and Frequency Similarity Based on GAN},
  author={Zhang, Junsan and Feng, Yeqi and Wang, Chao and Shao, Mingwen and Jiang, Yujie and Wang, Jian},
  journal={Neurocomputing},
  pages={126279},
  year={2023}
}