ITU Computer Vision Lab Rotated Mnist Contest

Introduction

To the attention of our new undergraduate students,

A competition has been organized by the ITU Computer Vision Laboratory for you. You are responsible for completing the tasks assigned to you on the MNIST data set in this contest, in which you will participate using Kaggle. The winners will receive rewards in addition to the experience gained by participating in this competition.

Competition Description

MNIST (Modified National Institute of Standards and Technology) is the de facto computer vision “hello world" dataset. This famous dataset of handwritten images has been used to test classification methods since its introduction in 1999. MNIST remains a dependable resource for researchers and learners alike as new machine learning approaches develop.

Your task is to accurately identify digits from a dataset of tens of thousands of handwritten photographs in this competition. We've put the rotated ones inside the original MNIST data to make things a little more complicated. We invite you to try out different algorithms to see what works best and how different strategies compare.

Practice Skills

  • Computer vision fundamentals including simple neural networks
  • Classification methods

Acknowledgments

http://yann.lecun.com/exdb/mnist/index.html has further information about the dataset, including the algorithms tried on it and their success rates.

Evaluation

The goal of this competition is to discover what a single handwritten number is and at what angle it is rotated by taking a picture of it. For every test set, you should predict the correct label and rotation angle.

Submission File Format

The file should contain a header and have the following format:

ImageId,Label,Angle
1, 1 , 𝛳
2, 8 , 𝛳
3, 5 , 𝛳
etc.

Prizes

Join The Competition