Skip to content

AniketGurav/RegCapsNet

 
 

Repository files navigation

RegCapsNet with ResNet - Keras/Tensorflow

This is a Keras and Tensorflow implementation of Regularized CapsNet with ResNet.

To know more about our proposed model, please refer to the original paper


Run "CapsNet.py" for the basic CapsuleNetwork model.


Run "RegCapsNet.py" for the Regularized version of CapsuleNetwork model.


To run our "RegCapsNet Conjugate with ResNet" follow the below steps:

1- To prepare the dataset, download it from the following and then run "Preparing.m"

2- Run "mainResNet.py" for the training model using ResNet-18. The best model will be saved in the folder "ResNetModels"

3- Run "ExatrctResNetFeatures.py" for extracting ResNet features. This code loads the best model from folder "ResNetModels" (in step 1) and then extracts train and test features in a specific layer number. Features will be saved in the folder "ResNetFeatures"

4- Run "RegResCapsNet.py" aiming signatures classification. This file used features of step 2 (which are saved in folder "ResNetFeatures") as input data.


Dataset

1- Cedar dataset is Available at (http://www.cedar.buffalo.edu/NIJ/data/signatures.rar) (a number of samples of original and pre-processed dataset are available in the Datasets folder).


For more information about loading dataset or setting the parameters, please refer to utilities folder.


Good luck

Cite information

Mahdi Jampour, Saeid Abbaasi, Malihe Javidi,

CapsNet Regularization and its Conjugation with ResNet for Signature Identification,

Pattern Recognition, Volume 120, 2021, 107851, https://doi.org/10.1016/j.patcog.2021.107851.

Paper is open access from the following link till August 29, 2021:

https://authors.elsevier.com/a/1dNsQ77nKcuF4

About

Regularized CapsNet Conjugate with ResNet

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 93.4%
  • MATLAB 6.6%