Chord Recognition in Beatles Songs

Chord Recognition in Beatles Songs

While a graduate student at MIT’s Media Lab, I collaborated with office-mate Vi­ctor Adán to explore how if we might train a machine to recognize chord changes in music. We tried multiple models to solve the problem, including Support Vector Machines, Neural Networks, Hidden Markov Models, and a few variations of Maximum Likelihood systems.

We chose Beatles tunes as a subset of the larger problem and trained our systems with 16 songs from three of their albums. Our systems processed 2700 training samples, 150 validation samples, and 246 testing samples. Our most successful system, a Support Vector Machine, achieved 68% accuracy in testing.

Our intention was to further the research which will lead to applications such as automatic transcription, live tracking for improvisation, and computer-assisted (synthetic) performers. Our models were an extension of the research provided by the following papers:

  • Musical Key Extraction from Audio, Steffen Pauws
  • Chord Segmentation and Recognition using EM-Trained Hidden Markov Models, Alexander Sheh and Daniel P.W. Ellis
  • SmartMusicKIOSK: Music Listening Station with Chorus-Search Function, Masataka Goto
  • A Chorus-Section Detecting Method for Musical Audio Signals, Masataka Goto

Main Website

Similar Posts

  • DoubleTalk

    Doubletalk, a two player audio-manipulation game was my first serious endeaver with the Gameboy. The game used the Pocketvoice, a Gameboy cartridge with a built-in amplified speaker and microphone. In Doubletalk, players record themselves, reverse their recordings, then try to guess what each other is saying.

  • SoundBlocks

    SoundBlocks is a tangible environment where youth connect blocks to describe network dataflow. The environment explores digital sound manipulation as a personal, meaningful and fun artistic endeavor, rather than as a venture into mathematical, electronic or networking relationships.

  • Still Life

    In 2011, as part of Hack.Art.Lab, I collaborated with composer Mary Ellen Childs and percussionist Michael Holland to create live animation triggered by live performance of Mary Ellen Childs’ composition “Still Life.” We analyzed the piece into 11 sections and created algorithmic video triggered by sound and motion to match each of the 11 sections. The video was projected…

  • LegalLanguage

    I wrote LegalLanguage, a scripting language for lawyers at Legal Services Corporation in West Virginia. The staff used LegalLanguage to write simple scripts that could then ask clients questions, give guidance, and print out the appropriate forms. This freed up resources to focus on the large number of cases involving domestic violence.

  • Contrapuntal Composer

    Contrapuntal Composer is Prolog code which writes music for three simultaneous voices. Depending on initial parameters, it can write a fugue, a rondo, or any other contrapuntal form. Contrapuntal Composer obeys the rules of good voice leading within each voice and between the voices.