This talk will run from 11:05am-12:00pm and be followed by another public lecture by Shlomo Dubnov on CIRMMT and the Digital Humanities from 3:30-4:30pm, also in room A832
ABSTRACT
The problem of motif discovery is important for a variety of musicological, music information retrieval and creative applications. Traditionally this problem was handled using symbolic music representations that allow pattern patching based on string search or geometric / metrical approaches. In the talk I will discuss a novel method for audio analysis that allows optimal symbolization of the audio recording by finding the most informative sound representation. Our current experiments show better results for finding repeating patterns in audio compared to other state of the art methods. Another advantage of the method is the generative nature of the sound representation that had been used for machine improvisation and query guided synthesis. Applications to data mining in massive sound and time-series databases will be discussed in the talk.
ABOUT SHLOMO DUBNOV
Shlomo Dubnov is a Professor in UCSD Music Department's Music Technology area. He graduated from the Jerusalem Music Academy in composition, and holds a doctorate in computer science and musicology from the Hebrew University, Jerusalem. Dubnov worked as a researcher at Institute for Research and Coordination of Acoustics and Music (IRCAM), in Paris, and headed the Multimedia track in Communication Systems Engineering in Ben-Gurion University. He is a Senior Member of IEEE and serves as a Director of the Center for Research in Entertainment and Learning at UCSD's Qualcomm research institute CALIT2.