DOI: 10.1080/09298210701653344
关键词: Speech recognition 、 Rhythm 、 Beat tracking 、 Dynamic programming 、 Beat detection 、 Audio signal 、 Beat (music) 、 Mathematics
摘要: Abstract Beat tracking – i.e. deriving from a music audio signal sequence of beat instants that might correspond to when human listener would tap his foot – involves satisfying two constraints. On the one hand, selected should generally moments in where is indicated, for instance by onset note played instruments. other set beats reflect locally-constant inter-beat-interval, since it this regular spacing between times defines musical rhythm. These dual constraints map neatly onto optimized dynamic programming, local match, and transition cost. We describe tracking system which first estimates global tempo, uses tempo construct cost function, then programming find best-scoring as well corresponding high ‘onset strength’ function derived audio. Th...