• No products in the cart.

Did you know that a survey said that 94% of pet owners say their animal pal makes them smile more than once a day? There are no prizes for guessing which are the most popular pets on the planet: they are cats and dogs. But how do you choose which one to adopt? How do you know if you’re a cat person or a dog person? And finally, how do you choose which breed to adopt (because let’s face it, they all look cute)? To make matters more difficult, there are around 340 breeds recognized by the Fédération Cynologique Internationale (FCI), the governing body of dog breeds, which is also called World Canine Organisation. On the other hand, The International Cat Association (TICA) recognizes 58 standardized breeds of cats.

So till you decide on which animal to adopt here is an interesting challenge for you. Can you make your computer identify cat and dog breeds with some machine learning magic? The challenge is two-fold. You have to train your machine to classify if the animal image given is of a cat or a dog but also to identify the particular breed. You will need to leverage the best and the most modern machine learning techniques out there.

We are looking for dedicated data scientists and machine learning engineers to write algorithms to identify the animal and the breed.


The dataset contains cat and dog images. The image dataset is based on 10 different breeds of cats and dogs combined. The train set consists of approx 600 images of each breed and the test set consists of 100 images approx. Each maximum dimension of the image is 400×400 and the minimum dimension is 100×100. The depth of every image is 3. The training set consists of 6,206 images covering 10 different breeds of cats and dogs combined. The test set consists of 1,096 images covering 10 different breeds of cats and dogs. Please use the sample submission for submitting a proper test result.

Download the data here and get started. Open the below link in your browser to automatically download all data.


Problem Statement

Based on the training data, predict if the image in test dataset is that of a cat or a dog and further predict the breed of the dog.

Hackathon Reviews


  • 5 stars0
  • 4 stars0
  • 3 stars0
  • 2 stars0
  • 1 stars0

No Reviews found for this hackathon.


© Analytics India Magazine