作者: Simon Dixon
关键词: Beat detection 、 Computer music 、 Speech recognition 、 Software system 、 Cluster analysis 、 Popular music 、 Digital audio 、 Time signature 、 Beat (music) 、 Computer science
摘要: Beat tracking is what people do when they tap their feet in time to music. We present a software system which performs this task, processing music standard digital audio format and estimating the locations of musical beats. A time-domain algorithm detects salient acoustic events, then clustering groups intervals between events obtain hypotheses about current tempo. Multiple competing agents track these throughout music, with further being created at decision points. The output for each agent sequence beat locations, evaluated its closeness fit data. This approach assumes no previous knowledge such as style, signature or approximate tempo; all required information derived from has been tested various styles (popular, jazz, classical) robustly, rarely making errors popular recovering quickly more complex despite fact that high level encoded system.