Haruto Takeda Takuya Nishimoto Shigeki Sagayama
Graduate School of Information Science and Technology,
The University of Tokyo
Hongo, Bunkyo-ku, Tokyo 113-8656 Japan /
{takeda,nishi,sagayama}@hil.t.u-tokyo.ac.jp
For rhythm recognition human-performed MIDI signals of polyphonic music,
we proposed an approach based on probabilistic models, in which Viterbi
search in HMM is used to estimate rhythm and EM Algorithm to estimate
tempo.
The results of experimental evaluation are also reported.