In this challenge we present Kvasir-Capsule, a large VCE dataset collected from examinations at Hospitals in Norway. Kvasir-Capsuleconsists of 118 videos which were used to extract a total of 4,820,739 image frames and around 40k+ of these images were verified medically and labelled.

The dataset can be downloaded from the given link: https://drive.google.com/file/d/1yDcNMTI-8Zq2Mvs8qe7pLNpPx4BW5Yta/view?usp=sharing

From the above link you will be able to download a zip file named ‘The Kvasir-Capsule.zip’. After you extract this zip file, you will get four files:

train - contains all the images that are to be used for training your model.
Train.csv - this csv file has contains the labels for the training images
test - contains 9000+ images. For these images you are required to make predictions
Test.csv - this is the order of the predictions for each image that is to be submitted on the platform. Make sure the predictions you submit are with their image’s filename in the same order as given in this file.

sample_submission.csv- this is the sample submission file for the data sprint

Saving Prediction File & Sample Submission

You can find more details on how to save a prediction file here: https://discuss.dphi.tech/t/how-to-submit-predictions/548

Sample submission: You should submit a CSV file with a header row and the sample submission can be found below:

label

Normal clean mucosa

Normal clean mucosa

Normal clean mucosa

Ileocecal valve

Normal clean mucosa

Normal clean mucosa

Normal clean mucosa

Normal clean mucosa

Foreign body

Normal clean mucosa

Normal clean mucosa

Pylorus

Pylorus

.

.

.

Etc.

Note that the header name should be ‘label’ else it will throw an evaluation error.

Notebook Submission

Steps to follow to submit your notebook:

Go to the Notebook section on this challenge page
Click on the 'Upload Notebook' button on top left corner as shown in the image below:
Make sure to make the visibility of your notebook as 'public' while uploading the notebook.
Provide Notebook Title, Summary, add some tags, go to next, chose a file and click on Upload.

Acknowledgement

Vestre Viken Health Trust (VV) in Norway

Data Sprint #38: The Kvasir-Capsule Dataset

Challenge Starts

Registration Ends

Challenge Ends