/ML-Model-Cartoon-or-Real

Primary LanguageJupyter Notebook

ML-Model-Cartoon-or-Real

Introduction

This contains a Machine Learning Model to decide whether a given image is an image of a cartoon or of a real person.

Pipeline

Dataset of various picture sizes is taken and split into two parts - 80% for training & rest 20% for testing.

Images are flattened and then reduced to 80 dimensions using Kernel Principle Component Analysis with 3 degree rbf kernel. Then, a multi layer perceptron classifier is used with 2 hidden layers & adam optimization.

Results

Obtained accuracy on testing data is 97.18% with a precision score of 1.0, a recall score of 0.95 & an F1 score of 0.97.

Running the code

Open the notebook model.ipynb & run all the cells in order. A brief report is also present with the name report.pdf.