Skip to main content

Table 7 Evaluation measures for onset detection on subsets of the AAM dataset. F-measure (F), precision (P), and recall (R) are provided for all three methods on all available dataset tracks (A), as well as on subsets of specific tempo segments (S: slow, M: medium, F: fast) and segments with and without drums (\(\textrm{D}\), \(\overline{\textrm{D}}\)). Best values are marked with bold font

From: AAM: a dataset of Artificial Audio Multitracks for diverse music information retrieval tasks

Measure

Kla99

Librosa

MIRtoolbox

\(F_{\textrm{A}}\)

0.224

0.787

0.515

\(P_{\textrm{A}}\)

0.228

0.834

0.473

\(R_{\textrm{A}}\)

0.226

0.752

0.582

\(F_{\textrm{S}}\)

0.105

0.821

0.424

\(P_{\textrm{S}}\)

0.095

0.861

0.355

\(R_{\textrm{S}}\)

0.122

0.796

0.552

\(F_{\textrm{M}}\)

0.196

0.793

0.506

\(P_{\textrm{M}}\)

0.192

0.841

0.459

\(R_{\textrm{M}}\)

0.205

0.762

0.577

\(F_{\textrm{F}}\)

0.333

0.742

0.599

\(P_{\textrm{F}}\)

0.369

0.806

0.596

\(R_{\textrm{F}}\)

0.311

0.702

0.611

\(F_{\textrm{D}}\)

0.222

0.804

0.517

\(P_{\textrm{D}}\)

0.223

0.830

0.480

\(R_{\textrm{D}}\)

0.225

0.782

0.575

\(F_\mathrm {\overline{D}}\)

0.228

0.673

0.522

\(P_\mathrm {\overline{D}}\)

0.258

0.851

0.473

\(R_\mathrm {\overline{D}}\)

0.218

0.603

0.625