Capsule-net for Urdu Digits Recognition

This repository contains code related to the publication "Capsule-net for Urdu digits recognition."

Dataset Preparation

Before running the code, make sure to:

• Split the dataset into two folders:
  • data_train
  • data_test

These folders should be placed in the root directory of the project and retain the same naming.

Running the Code

To run the model from the terminal, use one of the following commands:

python capsulenet.py

python capsulenet.py --epochs 50

python capsulenet.py --epochs 50 --num_routing 3

Cite

Iqbal, Talha, Hazrat Ali, Muhammad Muneeb Saad, Shahid Khan, and Camel Tanougast. "Capsule-net for Urdu digits recognition." In 2019 10th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), vol. 1, pp. 495–499. IEEE, 2019.

Disclaimer

The dataset is intended solely for academic research purposes. Proper acknowledgement of the authors is required if the dataset is used in any work or publication.