Implementing a gesture-based matchmaking system in an audio exercise game

• Main Author: Cruz, Ian Lazaruz
• Format: text
• Language: English
• Published: Animo Repository 2017
• Subjects: Computer games; Internet games; Video games; Electronic games
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/5804
• Institution: De La Salle University

Many games utilize matchmaking to improve the gaming experience for the players of the game. In this research, the use of skill-based versus style-based matchmaking is explored under the context of a Boxing Audio Exercise Game. Player actions, which are retrieved through the Kinect Sensor, are used to determine how each player plays the Boxing Audio Exercise Game. An 82.5% detection rate was achieved for the gesture recognition using the Hidden Markov Model. These gestures were then used to extract four clusters that describe the playing styles possible for the Shadow Boxing Game. The four clusters include players that focus primarily on attack (Aggressive Type), players that focus on determining the appropriate response to the opponents attack (Counter Punchers), players who beat all the objectives of the game (Completionists), and players who perform the least when compared with the other participants (Strugglers).