Litcius/Paper detail

Cricket Shot Detection using 2D CNN

John Bennilo Fernandes, Pandiri Sai Ram, Pandanaboina Madhu Varshith Yadav, Karur Pavan Kumar

202310 citationsDOI

Abstract

Cricket is a sport that is widely popular worldwide and played at both amateur and professional levels. Due to technological advances, it is now possible to use computer vision and deep learning techniques to automatically detect and classify cricket shots from videos and images. This research study intends to develop a system for the automatic detection and classification of cricket shots from images using create a 2D CNN with multiple layers namely convolution, pooling, flattening, and full connection. This model has achieved an overall accuracy of 91.5% in detecting different types of cricket shots. The proposed model is tested on a separate test set consisting of unseen shots and obtained similar results.

Topics & Concepts

CricketComputer scienceArtificial intelligenceConvolution (computer science)PoolingAmateurShot (pellet)Computer visionSet (abstract data type)Convolutional neural networkTest setPattern recognition (psychology)Artificial neural networkEcologyLawProgramming languagePolitical scienceOrganic chemistryBiologyChemistryVideo Analysis and SummarizationSports Analytics and PerformanceSports Dynamics and Biomechanics