Rythm Bot

• 

  Guitar Hero

Inspiration

The Inspiration came from wanting to make a machine to play a game

What it does

The game is streamed to the code via OBS and a small window is displayed for it. It then checks for colour inputs in this window and subsequently clicks the correct input with 95% accuracy

How we built it

We built it using python and OpenCV

Challenges we ran into

Getting the correct input speed and delay depending on spacing between the notes was very challenging

Accomplishments that we're proud of

We are proud that it achieves an accuracy of 95% and is better than 99% of humans attempting to play the game

What we learned

We learnt that frame rate is a major dependent on this model with higher frame rate causing higher accuracy of hits with more overhits. This can be applied to any other computer vision project too with more images displayed causing more input for the model

What's next for Rythm Bot

Applying the tool to various other rythm games

Built With

• opencv
• python