Image classifier

Image classifier is a supervised machine learning image classifier based on sklearn library and which uses Random Forest | k-nearest neighbor | Logistic Regression algorithms.

Setting environment using python

virtualenv env
source env/bin/activate
pip install -r requirements.txt

Running the project using python

python main.py --data-dir data_path --model randomForest

Setting environment using nix ( you should have nix installed )

nix-shell -I nixpkgs=channel:nixos-18.03

Running using nix

nix-env -if default.nix
image_classifier --data-dir data_path --model

--data-dir is the directory containing the data inputs/images. Images should have the format of classname.* ( for example car.1.png ).

--model is the machine learning algorithm to use. Three models are available { randomForest| knn | logisticRegression }

Making prediction

In python

python main.py --predict image.1.jpg

In nix

image_classifier --predict image.1.jpg