The application of AI in sports is widely researched as both coaches and players realise the significance of quantitative analysis that can be extracted from video matches. Detecting and segmenting in-play scenes in sport video
sequences is necessary in various applications such as quantitative game and performance analysis. In studies on video-based game and performance analysis of racket sports, much research efforts have been made to explore the relationships between predefined parameters and sports performance, and the methodologies for effective coaching.

Action recognition is one of the heavily studied topics in computer vision. Developing automatic methods for analyzing actions in videos are of particular importance for machine understanding of sports. For example, in badminton the recognition of jump smash, net play, defensive strokes and quality of service execution are amongst some research areas that are being studied. A different type of action recognition is applied for other racket sports such as in squash. There are several methods such as deep learning, transformers and generative adversarial networks will be explored in this research.