Practice #educational #classification #hard #cv

DGBRD

Dog breed classification

AIcrowd

1887

🛠 Contribute: Found a typo? Or any other change in the description that you would like to see? Please consider sending us a pull request in the public repo of the challenge here.

🕵️ Introduction

Dog, a man's best friend. Extremely friendly, cute and most of all, fluffy.

We've provided you a dataset of dog images. Your job is to classify them

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

💾 Dataset

The Stanford Dog Dataset contains over 20580 images of dogs from all over the world. There are also 120 classes of dogs. Initially this dataset was meant for object detection but we've converted it to an image classification problem

The dataset is split such that the train, test and validation sets have an equal distribution of the classes

📁 Files

Following files are available in the resources section:

train.zip - (17609 samples) This zip file contains 44609 various images corresponding to the training set.
train.csv - This csv file contains the image id and the corresponding label for the images in train.zip.
val.zip - (962 samples) This zip file contains 6204 various images corresponding to the validation set.
val.csv - This csv file contains the image id and the corresponding label for the images in val.zip.
test.zip - (2009 samples) This zip file contains 12551 various images and the test.csv file contains the filenames.

🚀 Submission

Prepare a CSV containing header as [filename,label] and predicted class as one of the 120 classes.
Name of the above file should be submission.csv.
Sample submission format available at sample_submission.csv in the resorces section.

Make your first submission here 🚀 !!