Music and Machine Learning: MIREX competition

The MIREX competition (Music Information Retrieval Evaluation eXchange) is an annual competition targeting challenging machine learning tasks related to music. You may choose one of the MIREX challenges as your Master's thesis project.

Some example projects are:

  • Genre classification of music audio files
  • Mood classification of music audio files
  • Audio fingerprinting
  • Pitch / melody tracking
  • Event detection in audio files
  • Beat tracking
  • Musical key detection
  • Cover song identification
  • Chord estimation 
  • Lyrics-to-audio alignminent
  • Drum transcription
  • Prediction of future musical notes
  • Tempo estimation

For a complete list of challenges to choose from, please visit the MIREX home page.

We will try to secure funding for successful students in order to travel to the annual ISMIR conference to present your results.


Suggested qualifications:

Machine learning / Deep learning (e.g. INF5860, INF4490),  Sound Analysis (MUS4831), Methods course in Cognitive Musicology (MUS4218)

Emneord: Music information retrieval, machine learning, classification, pattern recognition, prediction
Publisert 7. aug. 2018 14:17 - Sist endret 7. aug. 2018 14:24


Omfang (studiepoeng)