This week we finished our first prototype. We decided to focus on pattern recognition in this prototype. We designed the prototype to train boxers with combos memorization and variation. We will instruct them to finish a series of combos, teach them to repeat their hooks (3), and then take away the instructions and ask them to perform the combos correctly.
Technically wise, we faced with the problem of motion tracking. In the early test, we identified punches by detecting the hand velocities when the interaction happened. But when it was implemented into our first prototype, we found a problem. Because of the variety of boxing postures and hitting positions, punches might be incorrectly recognized. Thus, we chose to use collision detection to have a more accurate detection for now.
We presented our prototype to our expert, Francis Braza. He approved the general design for combos memorization and variation. He gave us instructions on switching our instruction method by making boxers focus more on the punching bag (ideally the opponent). Because of that, we take away the monitor screen and put all of our instructions on the punching bag.
We also prepared for the half presentation this week. We met several times with our instructors and polish our half presentation slides. During the preparation process, we discussed and decided our prototype iteration and our schedule for the following semester.
