Hey-Barrels

🕵️ Introduction

A popular method to tackle insomnia used to be counting sheep to fall asleep. While we are not trying to put you to sleep, we do have an AI twist of that practice! 

Can your AI model count pigs and barns from a given image?

🤔 Problem Statement

The task is to count the number of hay barrels and pigs in a given farm scenes. Along with the images of the farm scenes, segmentation images of hay barrels in the farm are also provided for each image in the training dataset. However no segmentation maps exist for those little pigs, so they are going to be a challenge to count. 

Understand with code! Here is getting started code for you.😄

💾 Dataset

The dataset consists of 1000 images of farms wherein each image there are some hay barrels and pigs present in the farm scene.

📁 Files

Instructions

Following files are available in the resources section:

• train.zip  - (500 samples) This zip contains two folders (images and segmentation_maps) where images correspond to images of the farm scenes and segmentation_maps contain corresponding segmented images of barrels. Along with this, there is meta-data.csv which contains barrels and pig counts under headers barrels_count, pigs_count respectively.
• test.zip - (500 samples) This zip file only contains images corresponding to farm scenes for which barrels and pig count have to be predicted.

🚀 Submission

• You need to submit a CSV file containing headers filename, barrels_count, pigs_count.
• Sample submission format available at sample_submission.csv in the resources section. Note keep the order of filename in submission.csv file same as that of given in  sample_submission.csv.

🖊 Evaluation Criteria

• The primary score will be counted as the accuracy of predicted barrel counts.
• The accuracy of predicted pigs count will be treated as a secondary score.

📱 Contact

• Avani Gupta
• Kanish Anand

📚 References

• Unity