symp

The dog breed prediction

  • Copy an url of the dog image into the input field (for example: https://oleg-ruban.net/img/portfolio/ml.jpg)
  • Submit data
  • Get the prediction in a separate tab

    • Note: The first request takes much longer to execute. It happens because if Lambda is not already running, AWS needs to deploy code and spin up a new container before the request can begin. After a cold start, the Lambda will remain instantiated for a while allowing any other call not to have to wait for this initialization to be done each time.



symp

  • Client

    Own

  • Date

    2020

  • Time Taken

    One Month

Info

This is my pet project in which I wanted to try my hand in Machine Learning.
I was interested to evaluate how a neural network will predict a breed of my loved pet that you could see on the preview for this page.

The Challenge

The main objective of the project is to create an application that is able to identify a breed of the dog by link on its image.

The Solution

  • By using the pre-trained architecture (ResNet - 34) I got to utilize a much deeper and more complex network architecture without the need to train such a deep network, which would require expensive hardware and lots of time.
  • This model was creatind by using fastai library and deployed on the Google Colaboratory platform that provides free GPU resources.
  • The deploying was done by hosting this model on Amazon S3 and implementing the serverless solution based on AWS Lambda and Amazon API Gateway.

Oleg Ruban