Nathan Chiu / Computer-Vision Guitar Tutor

Computer-Vision Guitar Tutor

Hack The Valley X

October 2025

Contributors: Sameer Bandha, Prisca Chien


As part of a 36-hour hackathon, we built Chordially, a computer vision–based tool that makes learning guitar chords simple, visual, and interactive.

How it Works

Fretboard Detection: Printed ArUcos placed around the fretboard of a guitar allow OpenCV to map the fretboard virtually.

Hand Tracking: MediaPipe (a framework by Google for building real-time perception pipelines using computer vision and machine learning) detects and tracks the user’s hands.

Finger Placement: Virtual fretboard overlays highlight the correct strings and frets for each chord, guiding users’ fingers into position and making chord learning more intuitive and visual.

Chord Matching: A database that links finger positions to chords allows the system to compare the user’s finger placement on the fretboard and generate an accuracy score.