This is an image classification program that predicts given images based on 10 classes listed below:
- Plane, Car, Bird, Cat, Deer, Dog, Frog, Horse, Ship and Truck It uses a Convolutional Neural Network to train on the training dataset, and predicts an ouput with over 70% accuracy. The cifar10 Dataset library from Keras was used to train the model of this application.
- Download and install Python 3 or higher from here
- Clone this repo to your local machine using:
git clone https://github.com/Tamer7/ImageClassifier.git
- Download the required packages by typing:
pip install -r requirements.txt
- Discard the Machine Learning model folder and use only the:
webapplication folder
- To run the django server follow these steps below:
a. cd into the webapplication folder (cd webapplication)
b. run the following command to start the server (python manage.py runserver)
c. Nagivate to the localhost shown in the terminal (in my case it is "http://127.0.0.1:8000/")
The Model used and developed is a CNN neural network, it was trained on 60,000 images, 50,000 training and 10,000 testing. 6,000 images where trained per class. The model was compiled using adam optimizer and was trained on 15 epcohs as the result shown below:
- Model training accuracy reached to : 80.2 percentage
- Model testing accuracy reached to : 70.06 percentage
- Python
- Tensorflow Library
- Django
- Cifar10 dataset from the keras library
- Tamer Algarmakany