Last week I gave a keynote talk entitled “Why Research Music” at the IndabaX Nigeria conference. Indaba (meaning “gathering” in Xhosa) is an organization that “develops knowledge and capacity in machine learning and artificial intelligence in individual countries across Africa.”
I was invited to give the talk by my friend and colleague Prof. Sakinat Oluwabukonla Folorunso from Olabisi Onabanjo University. We met through the Women in Music Information Retrieval group and recently started collaborating with her on research involving automatic music classification with Nigerian music genres. Sakinat was a primary organizer for the IndabaX conference and I was honored that she asked me to give a talk.
For my part, I have always been a little jealous of the large European or UK-based labs that focus on music information retrieval (MusicIR) research. In the US, we tend to have small labs with often just one professor and a couple of students. However, working with Sakinat, I realized just how good I had it in terms of being able to connect with a strong group of mentors, colloraborators, and students within my region (e.g., MusicIR researchers in the Northeastern US and Canada) and country.
So the goal for my talk to the Nigerian Machine Learning community was to promote music as a worthy domain of study. I also wanted to connect Sakinat and other researcher in Nigeria with other colleagues from the International Society for Music Information Retrieval (ISMIR) to help promote our research area. I encourage Sakinat to contact the ISMIR board and ask for volunteers to contribte to IndabaX.
In the end, we had three contributions:
- I gave an aspirational talk about the field of Music IR
- Meinard Müller gave a tutorial on content-based music analysis based on his excellent Fundamentals of Music Processing Book and Python code samples. (Meinard is the current ISMIR board president)
- Johanna Devaney represented both music IR and music cognition as a panelist on a Women in Machine Learning panel. (Johanna and I served as ISMIR scientific program chairs for the 2016 ISMIR conference in NYC.)
Our long term hope is create a home for music IR research in Nigeria as well as other countries in the region. As far as I know, we have never had a ISMIR paper (in 21 annual conferences) from a researcher based in Nigeria or Africa at large. This is something that needs to change since there are so many rich music traditions and talented musician from Nigeria.