2014:Audio Chord Estimation Results Billboard 2012

Introduction

This year, we have started a new evaluation battery for audio chord estimation. This page contains the results of these new evaluations for an abridged version of the Billboard dataset from McGill University, including a representative sample of American popular music from the 1950s through the 1990s, as used for MIREX 2012.

Why evaluate differently?

• Researchers interested in automatic chord estimation have been dissatisfied with the traditional evaluation techniques used for this task at MIREX.

• Numerous alternatives have been proposed in the literature (Harte, 2010; Mauch, 2010; Pauwels & Peeters, 2013).

• At ISMIR 2010 in Utrecht, a group discussed alternatives and developed the Utrecht Agreement for updating the task, but until this year, nobody had implemented any of the suggestions.

What’s new?

More precise recall estimation

• MIREX typically uses chord symbol recall (CSR) to estimate how well the predicted chords match the ground truth: the total duration of segments where the predictions match the ground truth divided by the total duration of the song.

• In previous years, MIREX has used an approximate CSR by sampling both the ground-truth and the automatic annotations every 10 ms.

• Following Harte (2010), we view the ground-truth and estimated annotations instead as continuous segmentations of the audio because (1) this is more precise and also (2) more computationally efficient.

• Moreover, because pieces of music come in a wide variety of lengths, we believe it is better to weight the CSR by the length of the song. This final number is referred to as the weighted chord symbol recall (WCSR).

Advanced chord vocabularies

• We computed WCSR with five different chord vocabulary mappings:

1. Chord root note only;
2. Major and minor;
3. Seventh chords;
4. Major and minor with inversions; and
5. Seventh chords with inversions.

• With the exception of no-chords, calculating the vocabulary mapping involves examining the root note, the bass note, and the relative interval structure of the chord labels.

• A mapping exists if both the root notes and bass notes match, and the structure of the output label is the largest possible subset of the input label given the vocabulary.

• For instance, in the major and minor case, G:7(#9) is mapped to G:maj because the interval set of G:maj, {1,3,5}, is a subset of the interval set of the G:7(#9), {1,3,5,b7,#9}. In the seventh-chord case, G:7(#9) is mapped to G:7 instead because the interval set of G:7 {1, 3, 5, b7} is also a subset of G:7(#9) but is larger than G:maj.

• Our recommendations are motivated by the frequencies of chord qualities in the Billboard corpus of American popular music (Burgoyne et al., 2011).

Most Frequent Chord Qualities in the Billboard Corpus

Quality	Freq.	Cum. Freq.
maj	52	52
min	13	65
7	10	75
min7	8	83
maj7	3	86

Evaluation of segmentation

• The chord transcription literature includes several other evaluation metrics, which mainly focus on the segmentation of the transcription.

• We propose to include the directional Hamming distance in the evaluation. The directional Hamming distance is calculated by finding for each annotated segment the maximally overlapping segment in the other annotation, and then summing the differences (Abdallah et al., 2005; Mauch, 2010).

• Depending on the order of application, the directional Hamming distance yields a measure of over- or under-segmentation. To keep the scaling consistent with WCSR values (1.0 is best and 0.0 is worst), we report 1 – over-segmentation and 1 – under-segmentation, as well as the harmonic mean of these values (cf. Harte, 2010).

Comparative Statistics

• coming soon...

Submissions

	Abstract	Contributors
CB3	PDF	Taemin Cho & Juan P. Bello
CB4	PDF	Taemin Cho & Juan P. Bello
CF2	PDF	Chris Cannam, Matthias Mauch, Matthew E. P. Davies, Simon Dixon, Christian Landone, Katy Noland, Mark Levy, Massimiliano Zanoni, Dan Stowell & Luís A. Figueira
KO1	PDF	Maksim Khadkevich & Maurizio Omologo
KO2	PDF	Maksim Khadkevich & Maurizio Omologo
NG1	PDF	Nikolay Glazyrin
NG2	PDF	Nikolay Glazyrin
NMSD1	PDF	Yizhao Ni, Matt Mcvicar, Raul Santos-Rodriguez & Tijl De Bie
NMSD2	PDF	Yizhao Ni, Matt Mcvicar, Raul Santos-Rodriguez & Tijl De Bie
PP3	PDF	Johan Pauwels & Geoffroy Peeters
PP4	PDF	Johan Pauwels & Geoffroy Peeters
SB8	PDF	Nikolaas Steenbergen & John Ashley Burgoyne

Results

Summary

All figures can be interpreted as percentages and range from 0 (worst) to 100 (best). The table is sorted on WCSR for the major-minor vocabulary. Algorithms that conducted training are marked with an asterisk; all others were submitted pre-trained.

Algorithm	Root	MajMin	MajMin + Inv	Sevenths	Sevenths + Inv	Mean Seg	Under-Seg	Over-Seg
CB3	79	77	75	65	63	86	84	89
CB4*	79	76	74	65	63	86	82	90
NMSD2	77	76	74	65	63	84	82	86
KO2*	78	76	74	62	60	85	82	89
NMSD1	78	76	74	64	62	84	82	86
KO1	77	76	74	58	56	85	83	87
PP3	74	73	70	53	51	84	84	84
CF2	74	72	70	55	53	84	85	83
NG1	73	71	69	52	50	84	82	86
PP4	73	70	68	55	53	84	82	87
NG2	69	67	65	48	47	83	83	83
SB8	9	6	6	5	5	52	92	36

download these results as csv

Comparative Statistics

• coming soon...

Complete Results

More detailed about the performance of the algorithms, including per-song performance and the breakdown of the WCSR calculations, is available from this archive:

• BillboardTest2012.zip

Algorithmic Output

The recognition output and the ground-truth files are available from this archive:

• BillboardTest2012Output.zip

We hope to generate a graphical comparison of all algorithms against the ground truth early in 2014.