Miguel Molina-Solana

One of the most challenging problems for the Sound and Music Computing field is the understanding of the way performers use expressive resources of a given instrument to communicate with the audience. Working directly with commercial recordings is a good opportunity for tackling this implicit knowledge. Nevertheless, the huge amount of information to be analyzed suggests the use of non-supervised techniques. These techniques have to deal with imprecise analysis and to manage the information in a more broader perspective. This work presents a new approach, Trend-based modeling, for identifying professional performers in commercial recordings. Departing from information automatically extracted by state-of-the-art extraction tools, our approach performs a qualitative analysis of the detected trends for a given set of different melodic patterns. We tested our system with a dataset of violin recordings from 23 different professional performers.

@incollection{Molina2009,
  title = {A Trend Based Model for Identifying Violinists},
  author = {Molina-Solana, Miguel},
  booktitle = {New Trends on Intelligent Systems and Soft Computing},
  publisher = {Editorial Universidad de Granada},
  year = {2009},
  address = {Granada, Spain},
  editor = {Cubero~Talavera, Juan~Carlos and Gonz\'alez~Mu\~noz, Antonio and Mar\'in~Ruiz, Nicol\'as},
  month = feb,
  pages = {143--160},
  volume = {2},
  timestamp = {104}
}