MachineHack is back again with another exciting hackathon for this weekend, and this time we take the data science enthusiasts to the past with the classic computer vision problem Dogs vs Cats.
Problem Statement & Description
In this hackathon, you will be provided with images of cats and dogs, and you must use your Computer Vision skills to build an image classifier to classify an image as that of a dog or of a cat. In this supervised image classification task, your goal is to classify the images into their respective classes using accuracy as a metric. The Dogs vs Cats is a classic dataset and has been used to train and evaluate models for binary classification tasks. With today’s state-of-the-art Computer Vision models, we expect all the participants to achieve an accuracy of more than 90%.
- Train (Folder): contains 9471 images of cats and dogs
- Test (Folder): contains 4059 images of cats and dogs
- Sample_Submission.csv: the format of submission accepted
- Train.csv: contains the file name and appropriate category for each image in the train data
- Test.csv: contains the file name for each image in the test data
To start quickly, we are providing this tutorial that has a few simple yet effective methods, which you can use to build a powerful image classifier using only a few training examples — just a few hundred or thousand pictures from each class you want to categorize. Building powerful image classification models using very little data
The datasets will be made available for download on July 10th, Friday at 6 pm IST
This hackathon and the bounty will expire on July 13th, Monday at 7 am IST
Sample Submission Format :
The top 3 competitors will receive a free pass to the Computer Vision DevCon 2020.
- One account per participant/team. Submissions from multiple accounts will lead to disqualification.
- The submission limit for the hackathon is 10 per day after which the submissions will not be accepted.
- All registered participants are eligible to compete in the hackathon.
- We ask that you respect the spirit of the competition and do not cheat.
- Use of any external dataset is prohibited, and doing so will lead to disqualification.
Hackathon Specific Rules
Participants must not manually label the images in submission. We work hard to host fair and fun contests and expect the same in return from the participants. However, we hold the right to wield the following measures:
- Spot check your code at any point in the competition.
- Disqualifying a participant on failure to provide proof of algorithms within a reasonable time frame
- Release a new test data at the end of the competition
- Access to your source code at the end of the competition to verify that the solution does not utilise any unfair means
- This hackathon will expire on 13th July, Monday at 7 am IST
The submission will be evaluated using the accuracy metric. One can use Sklearn’s accuracy_score to get a valid score.