TL;DR

This is an endless runner style game where the player moves their body in real life to avoid objects in-game. Their body position is tracked with Tensorflow and sent to a pygame-based app that translates their movements in real time. If they don't move in time and an object hits them, they lose. The longer someone stays alive, the higher their score.

Inspiration

We wanted to create a game that was fun to play and enjoyable, but at the same time didn't make players feel guilty that they were "wasting their time" on this game. Thus, we aimed to create a game that is educational and promotes healthy living. However, unlike many "educational" games out there, we wanted this one to actually be entertaining and competitive.

What it does

In the game, objects, which are modelled after common asian responsibilities or chores, come at the player's model. The player has to move their body in real life to avoid these objects. This includes moving side to side, jumping and ducking to dodge all types of obstacles. These movements are tracked through the player's camera and reflected in the game in real-time. The more objects they avoid and the longer they stay alive, the higher their score.

How it works

We used a tensorflow, AI-driven tracking model to precisely track the movements of 33 points on the human body through the player's front camera. We track these points and reflect the position of the them in-game by sending the location of these points to a pygame-based app that we created. The app takes in the coordinates of these points and moves the player seamlessly to reflect where the player is based on their tracking points.