Martial arts pose identification using deep learning

Conserving, supporting, and educating about martial arts is very important. Martial arts help us exercise and gift us with self-defense skills. As any martial art has various postures and poses of the body parts, extracting and recognizing complex human movements from videos is a challenge. In this project, the detection of various martial art poses with respect to Judo, Taekwondo, Karate, Boxing, Kickboxing, WWE, and many more will be achieved using CNN. The aim of this project is to identify and categorize the movements performed by the athletes using convolutional neural networks for training. In the beginning, the athlete’s bone structure is analyzed efficiently, and the action is judged accurately. Here, we shall use a special kind of Convolutional Neural Network (CNN) for analyzing the features of martial art and model to classify and tell us which martial art it is