research in artificial musicality
Hi, I'm Matthias. Welcome to my website and blog!
I'm a lecturer and researcher in the field of music informatics. I currently work as a Royal Academy of Engineering Research Fellow with the Centre for Digital Music at Queen Mary, University of London (see my Queen Mary web page).
Past work places include the Internet music platform Last.fm, where I worked as Research Fellow, the Japanese research centre AIST in Tsukuba, and, as a research student, the Centre for Digital Music. Find more info on my biography page.
My main research interest (and the subject of my PhD thesis) has been the automatic transcription of chords from audio, but I've also done work on segmentation, harpsichord tuning estimation and, recently, lyrics-to-audio alignment. Please do have a look at my publications website to learn more about my work, ask Google Scholar directly, or visit my Software site if you're more interested in just using it.
I just had the most pleasant radio interview with Sean Moncrieff of Newstalk Radio, Dublin. From what I can tell it’s a great, interesting show in general, endorsed by the Open University. I seem to have got my point across, too: we’ve got new votes streaming in on the Descent of Pop survey just now.
You can listen to the interview here.
Thanks to Sean and the Newstalk team!
I thought it might be high time to remind the world of the ground truth data (zip archive here) that my colleagues at AIST and I annotated, and which accompanies my paper on vocal/instrumental solo detection. On the AIST annotation page, there are, of course, many other interesting annotations, so have a look around there, too! The ground truth is provided for the whole RWC pop corpus (100 songs).
I’ve been involved in measuring the evolution of the charts for quite a while (at Last.fm, see the Anatomy of the Charts), and we’re opening a new chapter now. That new chapter needs your help! We want to dump those few annoying bad recordings from our database and learn about differences in audio quality in general.
Everything you need to know is on this website. But the real fun happens on the rating page.
It’s actually really interesting what crops up if you just randomly listen to the pop of the last 60 years, and if you do that for us, it would help us a lot.
So again, your help would be appreciated on this web page. You can also befriend us on Facebook: http://www.facebook.com/DescentOfPop
Today I had the pleasure to give a little seminar talk at the Technische Universität Dortmund. My slides are available here. In fact, I also had time to repeat my presentation of the drum pattern paper.
I’m currently at the ISMIR conference in Porto, Portugal. Very nice conference — and I’m hoping to contribute to that some more by presenting Sam Myer’s “Chordroid” real-time chord detection tool, which works on Android devices. In fact, I’ve actually organised a guitar. The demonstration is going to happen this afternoon in the infamous “Maus Habitos” bar, so the atmosphere should be friendly…
Something new, and, at least to me, exciting. And: it’s not about chords this time, can you believe it?
Accompanying web page is here.
Abstract: We present a corpus-based study of musical rhythm, based on a collection of 4.8 million bar-length drum patterns extracted from 48,176 pieces of symbolic music. Approaches to the analysis of rhythm in music information retrieval to date have focussed on low-level features for retrieval or on the detection of tempo, beats and drums in audio recordings. Musicological approaches are usually concerned with the description or implementation of manmade music theories. In this paper, we present a quantitative bottom-up approach to the study of rhythm that relies upon well-understoo… 
Our Songle.jp web service for interactive music analysis and annotation can now analyse and playback PIAPRO and SoundCloud content. This great new development from my old research group around Masataka Goto at AIST in Japan means that the number of interesting music tracks that are supported has multiplied. Previously the system relied on content that was stored in freely available and directly linked audio files, so in practice the content was mainly low quality … and usually also Japanese.
… or so I’m told, in a programme about DarwinTunes and other music stuff involving evolution, it’s called Darwin’s Tunes. Ah, well, it should be interesting anyway, so set your alarm clocks for next Wednesday at 9 pm (British Summer Time)!
