PyTorch Image Classifier

About

• As the final part of the Facebook AI scholarship challenge, I built a deep learning image classifier from scratch that uses computer vision to identify different species of flowers.

• The data set contains images of flowers from 102 different species curated by Oxford University. Which consist of a training set and a validation set. You can download the images from here or directly from the server using the links provided in the notebook. To test the model's accuracy using other images you will need to save the images on a test folder inside the flower_data folder.

• My final model was able to accurately classify the flowers 90% of the time. Below is an output of my model after you pass an image through it.

The project is broken down into multiple steps:

• Load and preprocess the image dataset.
• Train the image classifier on your dataset - train.py is a script that trains a deep neural network on the given data directory, with command-line options.
• Use the trained classifier to predict image content - predict.py is a script that predicts the flower name from an image, along with the probability of that name, given a trained classifier.

Dependencies

$ Python 3.5.2
$ pytorch = 0.4.1

Installation

$ git clone https://github.com/hassifow/PyTorch-Image-Classifier
$ cd PyTorch-Image-Classifier
$ git status

Author

• Hassif Abdulahi Mustafa