Cats & Dogs Classifier using Conventional Neural Network in Keras. You can get the dataset from here.

Following should be the directory structure. I've uploaded few images in the dataset folder so that you can understand more better the structure and file naming scheme.

project   
│
└───dataset
│   │
│   └───single_prediction
│   │   │   cat_or_dog1.jpg
│   │   │   cat_or_dog2.jpg
│   │
│   │
│   └───test_set
│   │   │
│   │   └───cats
│   │   │   cat.1.jpg
│   │   │   cat.2.jpg
│   │   │   ...
│   │   │
│   │   └───dogs
│   │       dog.1.jpg
│   │       dog.2.jpg
│   │       ...
│   │  
│   └───training_set
│       │
│       └───cats
│       │   cat.1.jpg
│       │   cat.2.jpg
│       │   ...
│       │
│       └───dogs
│           dog.1.jpg
│           dog.2.jpg
│           ...
│     
└───CATS_DOGS_CLASSIFIER.ipynb

• Python 3.6
• Keras 2.1.3
• Numpy
• Jupyter Notebook

You'll need (Keras)[https://keras.io/] installed to run the script. It can be installed using pip

pip install keras

We will also need numpy. It can also be installed using pip

pip install numpy

Download or Clone the repo and download the dataset from here and store it in the dataset folder using the structure showed above. After that, run either python script or the jupyter notebook. This CNN is trained using CPU only. If you are using GPU, feel free to add more layers or increase number of neurons.

Once the model is trained, you can use any image of dog or cat stored in single_prediction folder to predict either it is a dog or cat in the image.

1. Fork it
2. Create your feature branch: git checkout -b my-new-feature
3. Commit your changes: git commit -am 'Add some feature'
4. Push to the branch: git push origin my-new-feature
5. Submit a pull request

Muhammad Ali Zia

This project is licensed under the MIT License - see the LICENSE.md file for details