Dog Breed Classifier

In this assignment, we are provided with dataset of dogs containing images of 120 different breeds of dogs. Our job is to make a Deep learning model that will be able to classify 10 different breeds of dogs. Dogs - beagle, chihuahua, doberman, french_bulldog, golden_retriever, malamute, pug, saint_bernard, scottish_deerhound, tibetan_mastiff

This is an end-to-end Machine Learning project, deployed on Azure Web Services.

Installation

1. Clone this repo

git clone https://github.com/AshiishKarhade/Dog-Breed-Classifier-Assignment.git

2. Change the directory

cd Dog-Breed-Classifier-Assignment

3. Install the requirements

pip install -r requirements.txt

4. Run the app

python app.py

Model Interference and Execution

This project takes image as input in base64 string format.

The end-points can be accessed on https://dog-breeder.azurewebsites.net/predict (for limited period) or on https://localhost:5000 (if you are running locally)

The input format needs to be :

{

image : <base64 encoded image>

}

Expected response:

{

breed : <resulting label>

score : <prediction score of the above label >

}