Automatic Detection Of Cricket Umpire Gestures Using Artificial Intelligence

Cricket umpire gestures play a vital role in ensuring effective communication during matches, yet their automatic detection remains a significant challenge. This study introduces a novel approach using a Flower Pollination-tuned Adjustable Long Short-Term Memory (FP-ALSTM) model for cricket umpire gesture recognition. The model leverages advanced AI techniques to enhance both accuracy and robustness. Despite facing challenges such as dependence on high-quality video data and gesture variability, the proposed method demonstrates strong performance. A dataset gathered from Kaggle, consisting of 390 images, with 312 used for training and 78 for testing, was employed to cover a range of umpire gestures. The FP-ALSTM model achieves a validation accuracy of 95.47%, a train accuracy of 97.24%,and an F1 score of 97.25%. While these results are promising, the model's effectiveness in varying conditions remains an area for further exploration. Overall, the FP-ALSTM model aims to improve gesture detection accuracy, offering potential benefits for real-time decision-making and match analysis.