Future of music information retrieval
subject of upcoming talk

Imagine hearing a song on the radio, and wanting to know more. Its melody is ringing incessantly in your ears, but you don't know the title, the artist or who wrote it. Now imagine being able to sing or hum that song to a computer that can recognize melodies and instantly provide you with relevant information: MP3 versions of all known versions of the song, as well as its title, score, artist and composer.

Does a system like this currently exist? No. Will it? Definitely, with the help of people like Dr. J. Stephen Downie, a specialist in music information retrieval, and assistant professor at the University of Illinois at Urbana-Champaign.

UNC's School of Information and Library Science will host "Evaluation Methods for Music Information Retrieval (MIR) and Music Digital Library (MDL) Systems," a talk by Dr. Downie, to be held at noon Friday, Nov. 14, in Room 208 of Manning Hall.

Downie will discuss his multi-disciplinary project which brings together experts in the fields of supercomputing, music audio, music notation and music metadata to create an evaluation testbed for these systems.

For more information, please contact Megan Winget at winget@email.unc.edu