With the progress of our society, image analytics has become a vital role while predicting an accurate judgement in regard of a problem. Kshitij, IIT Kharagpur and Affine Analytics present an AI image classification challenge that will test your knowledge in solving the real world problem.
Design an algorithm to detect whether a bowler delivery is a legal or No-Ball delivery using the images of bowlers in action.
Data to be used:-
Training: Open-source database(Participants need to scrap their own database)
Sample: Sample 4-5 images has been shared to have an idea of validation set images
Participants should use the Google Collab GPU to train the model. Affine will not provide any GPU access.
- Model pickle file (to be executed for the validation
- Well commented End to End code file (py or ipynb file)
- Training and Test result summary-
• Method / Algorithms used
• Final Hyperparameters
• Accuracy, Precision, RecallAffine will evaluate the accuracy based on Affine owned test images.
NOTE: The website link for the model submission has been shared where participants can submit using the provided link before 18th January 6 PM.
Rules and Regulations :
1. Maximum team size 3 members ( Members may be of different departments).
2. All participants must carry ID proof of their own.
3. All participants are required to bring the registration receipt with them.
4. Participants must be present during the event at the specified time. The absent team will be automatically disqualified.
5. In case of any inconvenience or doubts during the competition, participants are required to seek the help of the volunteers present or to the event coordinator.
6. In case of any disrupts, the final world will remain with the organization.
7. The decision of the Judges is final.