Warning!

Greek content is currently not available

Track popularity

The aim of track popularity or Hit Song Science is to apply machine learning techniques in order to capture some information from musical data that would explain the popularity of the respective musical tracks.

More
Track popularity

Audio content-based music similarity

Technologies that can find similar songs with respect to timbre, rhythm and harmony based solely on music audio. These can be applied to either music recommendation, automatic playlist generation or even cross-genre similarity estimation for musicological studies.

More
Audio content-based music similarity

Rhythm analysis

By computing intermediate & higher level features, a number of methods are developed that automatically recognize various aspects of rhythm, such as dance style, tempo, beat positions and time signature.

More
Rhythm analysis

Towards an Invertible Rhythm Representation

This work investigates the development of a rhythm representation of music audio signals, that (i) is able to tackle rhythm related tasks and, (ii) is invertible, i.e. is suitable to reconstruct audio from it with the corresponding rhythm content being preserved.

More
Towards an Invertible Rhythm Representation

Percussive & harmonic content separation

A very simple and intuitive method for separating the audio signal into two components: the percussive instruments and the non-percussive (or harmonic) instruments.

More
Separation of Percussive & Harmonic Contents of music audio