Screen shot from the video generated by the system.
This is a project made for the course Music Informatics. A system was created in Python, were the user input a Beatles song audio file, which then generates a video file of visualisations based on the musical content of that particular audio. Using a dataset of the songs The Beatles, a machine learning model was trained to recognise chords played in any given Beatles song, which maps different chords to different colors. The beat of the songs is mapped to the brightness of the image. The musical structure of the songs is mapped to changing to different images of The Beatles. My main responsibility was developing a Hidden Markov Model for the chord recognition. Below is a video of the results.
A video of from The Beatles music visualisation system.