One-Shot Layer-Wise Accuracy Approximation for Layer Pruning

This is a demo for the proposed method on VGG19_bn cifar100 to generate Figure 3 in the paper.

Setup requirements

# Virtual environment creation
virtualenv .envpy36 -p python3.6
source .envpy36/bin/activate
#Install libraries
pip install -r req.txt

pretrained weights

Download pretrained weights for CIFAR100 vgg19_bn from here

Run

python imprint_cifar.py -d cifar100 --arch vgg19_bn --pretrained PATH_TO_MODEL -c cifar100_vgg

Cite

If you find this code useful in your research, please consider citing:

@inproceedings{elkerdawy2020one,
  title={One-Shot Layer-Wise Accuracy Approximation For Layer Pruning},
  author={Elkerdawy, Sara and Elhoushi, Mostafa and Singh, Abhineet and Zhang, Hong and Ray, Nilanjan},
  booktitle={2020 IEEE International Conference on Image Processing (ICIP)},
  pages={2940--2944},
  year={2020},
  organization={IEEE}
}