π¨ Update in Test Dataset | π₯ Reminder to Download
π΅οΈ Introduction
Our mission to Mars is about to start πΈ but before we can take off we must resolve a mix-up.
Someone has mishandled the labeling of two Mars rover projects β Curiosity and Perseverance β we must classify them correctly.
Click here to access the starter-kit and avert the emergency!

πΎ Dataset
The given dataset contains images of two different rovers i.e. Curiosity and Perseverance of size 265*256 in `jpg` format. The images in train.zip and val.zip have their labels i.e. which rover it is in train.csv and val.csv. The labels for the images in test.zip needs to be predicted.
π Files
Following files are available in the resources section:
train.zip- (40000samples) This zip file contains rover images with images name corresponding toImageIDcolumn oftrain.csvtrain.csv- (40000samples) This csv file contains theImageIDcolumn corresponding totrain.zipandlabelcolumn as the name of rover.val.zip- (4000samples) This zip file contains rover images with images name corresponding toImageIDcolumn ofval.csvval.csv- (4000samples) This csv file contains theImageIDcolumn corresponding toval.zipandlabelcolumn as the name of rover.test.zip- (10000samples) This zip file contains rover images which will be used to evaluate the performance of the model.
π Submission
- Prepare a CSV containing
ImageIDcolumn corresponding totest.zipandlabelcolumn as rover name. - The name of the above file should be submission.csv.
- Sample submission format available at sample_submission.csv in the resources section.
Make your first submission here π !!
π Evaluation Criteria
During evaluation F1 score is used as Primary Score and Accuracy Score as Secondary Score will be used to test the efficiency of the model.
π Links
- πͺ Challenge Page: https://www.aicrowd.com/challenges/rover-classification
- π£οΈ Discussion Forum: https://www.aicrowd.com/challenges/rover-classification/discussion
- π Leaderboard: https://www.aicrowd.com/challenges/rover-classification/leaderboards
Sign Up with Email