Europeana Sounds at the Vienna Waves Music Hackday: the results

On 1 October, Europeana Sounds co-organised a full hackday as part of the Vienna Waves Music Festival. During this day we challenged programmers, developers, designers, musicians, and anybody else with an interest in music and technology to come and work on their own idea or on one of the proposed challenges.

On Friday evening, the organisers and participants came together in the Werkstätten- und Kulturhaus (WUK) to listen to presentations on different tools and challenges that they could work on. A mix of hardware, software and data was made available for the participants.

Joris Pekel brought the Europeana Music Collections dataset to the event and proposed the Europeana Sounds Genre Detection Challenge. In short, for a user of Europeana Music, it is useful to be able to search for particular music genres (e.g., free jazz, Irish folk, Baroque) to find what they are looking for in this vast amount of material, currently over 200.000 items. However, this information is not always available in the data. Currently, only about a fifth of the Europeana Music Collection has been labelled with a unified genre description. The participants were challenged to come up with solutions to detect genres in the data and music files.

The results

As a result of the hackathon, the Europeana Music dataset has been analysed in great detail. A lot of different aspects such as tempo, timbre and loudness have been processed so that the individual characteristics can be analysed. All these results have been made available on a special page set up for the hackday and are available for download for anyone who wants to work with it. However, the day was too short for the hackday participants to come up with a working solution to automatically detect genres. But, we laid great groundwork for further research and work.

The Europeana Music Collection was of great interest, as it is openly available and ready for use by software developers, and many useful connections have been made. Especially the work done by the company MusiMap, also present at the hackathon, shows great potential for further collaboration. MusiMap builds software to analyse music and index its characteristics.

In the end 16 hacks were submitted to the judges for review, varying from artificial intelligence bots that can help you find the right music that complements your mood, to devices that can be attached to the body during a live performance to influence the music depending on your position. All results can be seen in the video report created by Maarten Brinkerink of the Netherlands Institute for Sound and Vision (Nederlands Instituut voor Beeld en Geluid).

This hackday shows that the dataset shows great potential, but in a hackday of 8 hours only the first steps can be taken. We are therefore continuing our quest to look for better ways to automatically detect genres in music.

If you are interested in working with us, please get in touch!