Exploring Team Passing and Movement in Sports: A Deep Learning Framework

Sports are a type of competitive physical activity that aims to promote the actual abilities and capacity of a person based on team and player performance. The main agenda of a team in sports is to achieve a better score when compared to their opponent team. During this play period, the defensive players further try to predict the upcoming moves of the attackers to prevent them from scoring goals and vice versa. Such action is an outstanding illustration of a predictable nature and is a key part of human intelligence. In this way, the consistently developing interests in sports have paved the way for the development of various analytical methods and techniques to understand the boundaries of sports. Recent advances in deep learning algorithms have shown a way for many effective and successful applications. This proposed work aims to devise a novel system that is an integration of classification and CNN-based DenseNet framework for predicting the next move of a player based on the player’s attitude and body movements in a sport. The proposed model showed better accuracy compared to other classification models. Similarly, numerous existing problems in the field of mining, metal and fuel industries can be addressed with the support of advanced technologies like AI and CV. The CNN-based models and computer vision techniques can be used in mining industries to forecast blast-induced air-blast overpressure in open pit mines and also to distinguish the source area of microseismic occasions based on recorded ground tremors, therefore preventing mine catastrophe. Nevertheless, not only on-site but also the usage of multiple deep learning hybrid and detection algorithms helps in reducing irregularities and faults at manufacturing units with an increase in automation process and smart manufacturing techniques.

View article page

Download PDF