Exploring Transfer-Learning with Voting Concept for Identification of Cricket Event

Event detection is crucial in the modern cricket, as it accurately identifies key moments such as six, out and no ball in real-time. It enhances decision-making for coaches, analysts, and umpires, leading to better strategic planning and fair play enforcement. Additionally, it enhances fan engagement by offering instant replays and live updates while also serving as a valuable tool for summarizing matches. By detecting key events automatically, highlights can be generated without manual editing, making the process more efficient and accurate. In this article, the author(s) present a deep learning-based cricket event detection method based on transfer learning and voting concepts. The proposed approach includes integrating a well-known model like VGG16, VGG19, and ResNet50 and predicting the event based on voting results by analyzing the umpire action. The proposed method gives 94.69% accuracy in classifying six major cricket events, i.e., six, out, wide, no ball, drs, and no action.