We use the THETIS dataset for our project. http://thetis.image.ece.ntua.gr/

Sofia Gourgari, Georgios Goudelis, Konstantinos Karpouzis and Stefanos Kollias. THETIS: THree Dimensional Tennis Shots A human action dataset. In CVPR, International workshop on Behavior Analysis in Games and modern Sensing devices, June 2013

Our model files are in the "Code" directory. We extract frames from the videos, and try out 4 models - a base CNN with all 12 class, a CNN with 2 classes, changing our inputs from 3D to 4D and trying out the same CNN with 2 classes, and finally LRCNN. The last model has been referenced from the existing work done by Vincent Chow and Ohi Dibua. The link for their repository is as below: https://github.com/chow-vincent/tennis_action_recognition