Exploiting global features for tempo octave correction

Abstract

Tempo estimation is a fundamental problem in music information retrieval. Most approaches attempt to solve two problems: first finding a dominant pulse and second correcting the metrical level of this pulse. The latter has also been dubbed fixing the octave error. We propose an algorithm for tempo estimation that addresses both problems mostly independently. While using a standard pulse detection technique, for octave error correction, we exploit a simple relationship between a single global feature, average spectral novelty, and listener perception of musical tempo. The proposed method is extremely simple. Nevertheless, it outperforms most existing tempo estimation methods and is on par with the best-performing ones. It thus exemplifies that a global feature-based approach can significantly improve tempo estimation.

Contemporary music genre rhythm generation with machine learning

“…We are interested on modeling and playing with contemporary music genre rhythms such as footwork, trap, 2-step, gqom, drum and bass, and dembow. The appeal of these rhythms comes from their tempo octaves-that is, their actual tempo is ambiguous and can be perceived or embodied at one specific tempo, or half or double that tempo (Schreiber & Müller, 2014). For example, when dancing to a drum and bass track, people can embody the music as if the tempo is 160 beats per minute (bpm) or as if it is 80 bpm.…”

Bridging Music Using Sound-Effect Insertion

“…6 Worse yet, most beat-detection algorithms have a common issue with tempo octave errors. 14 To handle this problem, we adopt a technique of dual-tempo adjustment, which is similar to what Hiromi Ishizaki and his colleagues proposed. 3 That is, we will match the IBIs' tempo values of clip a-their double/half or their quadruple/quarter-to those of clip b.…”