OCAP: On-device Class-Aware Pruning for personalized edge DNN models

This repository contains the code for OCAP.

OCAP proposes a new class-aware pruning method based on the intermediate activation of input images to identify and remove the class-irrelevant channels.

For other details, please refer to the paper.

All the model definitions are available under ./models, and the auxiliary functions are available under ./utils.

OCAP on CIFAR

MobileNetV2_Cifar_Pruning.py

ResNet_Cifar_Pruning.py

Vgg_Cifar_Pruning.py

OCAP on ImageNet

Vgg_ImageNet_Pruning.py