Cat vs Dog / Classification of various dog breeds

An Image Classifier Project

This is a summary of my project for the Winter Internship Program (WIP 3.0) under the CSI chapter of LNMIIT, Jaipur.

This project has 2 phases. Each of them involves using an image classifier to acheive desired results using image data.

Phase 1

Aims to classify an input image as either a dog or a cat image.

Phase 2

Aims to classify images of dogs according to their various breeds.

The image input will be analyzed and the predicted result will be given as output. Machine learning algorithm (Convolutional Neural Networks) is used to classify the image. Various other approcahes/algorithms will also be tested along and their accuracies compared.

Tools used

1. python3
2. pandas
3. scikit-learn
4. tensorflow
5. matplotlib

Procedure

A breakdown of the approach I shall be using.

1. Finding the right dataset.

Details of the dataset will be updated.

2. Choosing the right models to run on.

I shall mainly be using Convolutional neural networks to build my model, along with other better approaches I might come across.

3. Training the model.

I shall use either my personal computer or cloud services, whichever seems more feasible for training the model.

4.Testing out.

5. Publishing results.

Authors

Made by Aasim Mukhtar

B.Tech, Mechanical Engineering

The LNM Institute of Information Technology, Jaipur