Homework Run

Homework Run is an augmented reality game that challenges players to stay active while avoiding obstacles and answering trivia questions. Using their body movements, players control their in-game character, dodging giant cellphones, hourglasses, and Instagram icons. If they collide with an obstacle, they must answer a true-false question on a topic of their choice. The game tracks the player's speed and detects when they're running, making it a fun and engaging way to stay active.

Inspiration

With the oncoming Subway Surfers brainrot of the new generation, the goldfish attention span, and inability to touch grass, we developed a game that forces the new generation (and us) to stay active in the only way we know: Subway Surfers.

As students, school can be hard. We were inspired by our struggles and decided to make it student themed!

How we built it

We used a combination of computer vision and machine learning technologies to bring Homework Run to life. Here's a breakdown of our tech stack:

OpenCV: We used OpenCV to capture and process video frames from the player's webcam, detecting their body movements and tracking their speed.

MediaPipe: We leveraged MediaPipe to detect the player's pose and movements, allowing us to track their running speed and detect collisions with obstacles.

OpenAI's GPT-3.5: We used OpenAI's GPT-3.5 language model to generate true-false questions on a topic of the player's choice, adding an educational element to the game.

Arduino and PySerial: We used Arduino to control LED lights that provide feedback to the player, and PySerial to communicate with the Arduino board.

Challenges we ran into

Procrastinating by wandering a mall for 5 hours. Also had some trouble detecting whether the player was running. Working with mediapipe was also hard - learned a lot about human anatomy.