Tools
Tools
Kanagal Sathyanarayana, Rohan (2023) CricNet: A Deep Learning Network to Enhance the Cricketing Analysis. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (1MB) | Preview |
Abstract
The present research examines the transformative effect using data and artificial intelligence perception upon strategy and training in cricket sport. CricNet, the idea and the primary objective that has been put forward, utilizes cutting-edge technologies involving convolutional neural network and transfer learning approaches to assess simultaneous streaming video as well as still images. VGG19, ResNet50, InceptionV3, MobileNet, EfficientNetB0, DenseNet121, Xception were adopted wherein InceptionV3 accomplished an amazing accuracy rate of 99.58%. This deep learning-based outcome classifies cricket shots in real time, integrating gesture recognition employing Detectron mode to derive reliable angles of the body. All these perspectives are subsequently displayed to coaches and players to examine, enabling useful information towards enhancement of skills. CricNet not just exposes specific players’ weakness but additionally allows coaches to take decisions based on data, which leads to improved training strategies. On top of that the technology facilitates in-game analysis permitting players as well as coaches to carry out strategic moves and modify their tactics all through the game.
Actions (login required)
 |
View Item |