Music/Signal Processing at the University of Tokyo, par Shigeki Sagayama, 09/09/2012

Vendredi 9 septembre à 11h
Télécom ParisTech (Dareau)
Salle: DA-006

Auteur: Pr. Shigeki Sagayama, Graduate School of Information Science and Technology, The University of Tokyo

Résumé:
Selected topics from multi-pitch analysis and source separation/modification of music signals based on GMM and NMF, key/tempo modification, chord detection, key identification, score following, genre/mood classification, rhythm recognition, vocal cancelation (producing a “Karaoke”), automatic accompaniment, automatic music composition, arrangement, fingering decision and other topics will be addressed in this talk. In the past decade, music signal and information processing has made a significant progress with probabilistic modeling and machine-learning algorithms such as hidden Markov model (HMM), Gaussian mixture model (GMM), non-negative matrix factorization (NMF) and conditional random field (CRF). A short introduction to the university of Tokyo will be also presented.

A propos de l’auteur:
Shigeki Sagayama is a professor at the University of Tokyo, Japan. After working long in speech recognition area at NTT and ATR Laboratories since 1974, he started his research on music signal and information processing in 1998 and now is the author or coauthor of 200 technical papers on music processing in addition to more than 500 papers in other areas. He is a fellow of IEICEJ and Chair, IEEE SPS Japan Chapter.

Comments are closed.