/django-image-classifier

Web-application that classifies given image by using Convolutional Neural Network

Primary LanguagePythonMIT LicenseMIT

Image Classifier

Web-application written in Django

Web-application that classifies given image by using Convolutional Neural Network. Neural Network model was trained using one of the commonly used dataset to train machine learning and computer vision algorithms – CIFAR-10.

Here are some numbers about training process:

  • Amount of epochs: 350
  • Training time: ~10 hrs
  • Accuracy: 94.5%

Installation

To get started with the application you need to complete following steps:

  • Install dependencies:
$ pip install -r requirements.txt
  • Create .env file with the following properties and set the appropriate values:
SECRET_KEY=
  • Run application:
$ python3 manage.py runserver 8000

Usage

After you run application you will be able to upload any image you want to classify and get the result of the classification by trained model.