The growing use of technology to permit more accurate and timely in-game decision-making enhances match adjudication and informs fans and the audience in professional sports competitions. The use of optical motion capture systems and video analytics has become the official support for major sports tournaments helping umpires to determine the penalty situations that are difficult to define by the human eye in fast-paced game events. In cricket, a front-foot-no-ball or, or overstepping, is a growing talking point among players and fans. There have been many occasions when this action has not been detected or called by the umpires. While some of such events make big stories many others remain only on the broadcast footage. Currently, the ICC mandates for international games is careful monitoring of every delivery that is subject to a dismissal by the TV umpire and the tech team stopping the game and rewinding video feeds for the slow motion of the player action. This can take many minutes of game time. Another illegal bowling action is defined as when the bowling arm is extended more than fifteen degrees before releasing the ball. This is called overextension. Due to the complex bio-mechanical movement of the bowling arm, it is almost impossible for the TV or on-field umpire to call this type of illegal bowling action, particularly because it has very heavy consequences for the player.In this paper, we present a computer vision software solution that continually tracks the key points in a bowler's body for illegal actions of overstepping and overextension. The system utilizes proprietary models that run on the run-out camera footage around the pitch and detect possible illegal actions almost in real-time. The software aims at simplifying the many tasks of an umpire and making adjudicating the match more efficient and more accurate, while no extra time is taken from the game. Our system has been endorsed by ICC after successful testing in the T20 cricket world cup 2022.

We are in the era that sport is increasingly defined by data. Rich data is a powerful enabling foundation for novel insights on the game and on player actions, and consequently used for fan engagement and better decision making. Detailed data that has specifications of each shot is currently missing in tennis, as the game dynamics is fast and it is beyond human ability to manually record all specifications of each and every shot, even using scoring software. In this paper, we present an intelligent system to automatically recognize player actions, ball and player movements, and important game events. The system annotates the video with a suitable set of keywords for fast retrieval in broadcast production. Various techniques of computer vision, alignment, filtering, and pre-trained deep learning models are utilized by our system. The evaluation of our results on multiple broadcast videos show great accuracy and timeliness. The implications of the work presented in this paper are profound in the current workflow in broadcast coverage of a tennis match where normally multiple video operators and judges are needed to identify events and retrieve the related clips from multiple cameras.

This work presents an early stage big data application, NetClips, for automation in broadcast production in the area of content management, delivery, and consumption. NetClips acts as a cognitive engine to automatically annotate video clips with contextual sports content during match production and on archival contents. Ensemble of deep learning architectures and advanced machine vision techniques are used in NetClips to accurately recognize players, objects, actions, and game events in every clip created during a sport broadcast. This application also has a great impact on production workflow, as it will lift up the burden of concurrent tagging from operators list of duties.