Thibault LangloisThibault Langlois has a Computer Science Engineering degree and a PhD in Computer Science on Reinforcement Learning from the Université de Technologie de Compiègne (France). He joined the Neural Networks at INESC in 1994 and teached at the Instituto Superior Técnico. Since 2001 he is Assistant Professor in the Informatics Department at the Science Faculty of Lisbon University. He is member of the Large Scale Informatics Systems Laboratory where he is doing his research activities on Machine Learning and its application to multimedia data mining. His current research topics include Video and Music Information Retrieval.
Thibault Langlois, HCIM/Faculdade de Ciências da Universidade de Lisboa
Gonçalo Marques, HCIM
Gonçalo Marques
Gonçalo MarquesGonçalo MarquesGonçalo Marques received a B.S.E.E. degree and a M.S.E.E. degree in electrical engineering from San Diego State University. He currently pursuing the Ph.D. degree from the Science Faculty of the Lisbon University. His research interests include signal processing, machine learning and music information retrieval.
We present a method for music classification based solely on
the audio contents of the music signal. More specifically, the audio
signal is converted into a compact symbolic representation that retains
timbral characteristics and accounts for the temporal structure of a
music piece. Models that capture the temporal dependencies observed in
the symbolic sequences of a set of music pieces are built using a
statistical language modeling approach. The proposed method is
evaluated on two classification tasks (Music Genre classification and
Artist Identification) using publicly available datasets. Finally, a
distance measure between music pieces is derived from the method and
examples of playlists generated using this distance are given. The
proposed method is compared with two alternative approaches which
include the use of Hidden Markov Models and a classification scheme
that ignores the temporal structure of the sequences of symbols. In
both cases the proposed approach outperforms the alternatives.
