Badminton Smashing Recognition through Video Performance by using Deep Learning

Nowadays, badminton become the hot trends sport in Malaysia due to influence of Lee Zii Jia which is Malaysian player and he has been participate men’s single Tokyo 2020 Olympic Game at Musashino Forest Sports Plaza Tokyo. Due this reason, analysis one major contribution analysing improving performance athlete. Hence, project constructs a smashing recognition through video by using deep learning. The main purpose evaluate models classifying types badminton. will be trained Deep Learning ResNet-18, GoogleNet VGG-16 best precision accuracy were compared. In project, we found that ResNet-18 97.51% 98.86% on both training testing datasets respectively software Jupyter. On other hand, highest 83.04% 97.20% hardware Jetson Nano.