BREVE: An HMPerceptron-Based Chord Recognition System

Abstract

Tonal harmony analysis is a sophisticated task. It combines general knowledge with contextual cues, and it is concerned with faceted and evolving objects such as musical language, execution style and taste. We present BREVE, a system for performing a particular kind of harmony analysis, chord recognition: music is encoded as a sequence of sounding events and the system should assing the appropriate chord label to each event. The solution proposed to the problem relies on a conditional model, where domain knowledge is encoded in the form of Boolean features. BREVE exploits the recently proposed algorithm CarpeDiem to obtain significant computational gains in solving the optimization problem underlying the classification process. The implemented system has been validated on a corpus of chorales from J.S. Bach: we report and discuss the learnt weights, point out the committed errors, and elaborate on the correlation between errors and growth in the classification times in places where the music is less clearly asserted.

Improving Audio Chord Estimation by Alignment and Integration of Crowd-Sourced Symbolic Music

“…Early works (Winograd, 1968;Maxwell, 1992;Temperley and Sleator, 1999) are grammar-or rule-based approaches to perform automatic tonal (Roman numeral) analysis. Other systems combine segmentation and template matching, applying tie-breaking rules (Pardo and Birmingham, 2002;Scholz and Ramalho, 2008); use probabilistic models to perform functional harmonic analysis on MIDI data (Raphael and Stoddard, 2004); or perform chord estimation using an HMPerceptron model, in which the domain knowledge is modelled in Boolean features (Radicioni and Esposito, 2010). In more recent research by Masada and Bunescu (2019), a model based on semi-Markov Conditional Random Fields (semi-CRFs) performs a joint segmentation and labelling of symbolic music.…”

Chord Recognition in Symbolic Music: A Segmental CRF Model, Segment-Level Features, and Comparative Evaluations on Classical and Popular Music

“…The performance measures were computed by averaging the results from the 10 test folds for each of the fold sets. Table 3 shows the averaged event-level and segment-level performance of the semi-CRF model, together with two versions of the HMPerceptron: HMPerceptron 1 , for which we do enharmonic normalization both on training and test data, similar to the normalization done for semi-CRF; and HMPerceptron 2 , which is the original system from (Radicioni and Esposito, 2010) Table 3: Comparative results (%) and standard deviations on the BaCh dataset, using Event-level accuracy (Acc E ) and Segment-level precision (P S ), recall (R S ), and F-measure (F S ).…”

Automatic Music Completion Based on Joint Optimization of Harmony Progression and Voicing

“…Still, the size of data sets of four-part voicings and harmony progressions is quite limited. In particular, the largest data set available to us that contains information on both harmony and voicing comprises only about 40 four-part chorales with harmony annotation [24]. We also had access to around 400 fourpart chorales without harmony annotation [25], and harmony annotation of around 80 classical pieces, but without four-part voicing [26].…”