• Corpus ID: 13456193

Towards Machine Learning of Expressive Microtiming in Brazilian Drumming

  title={Towards Machine Learning of Expressive Microtiming in Brazilian Drumming},
  author={Matthew Wright and Edgar Berdahl},
We have used supervised machine learning to apply microtiming to music specified only in terms of quantized note times for a variety of percussion instruments. The output of the regression schemes we tried is simply the microtiming deviation to apply to each note. In particular, we trained Locally Weighted Linear Regression / KNearest-Neighbors (LWLR/KNN), Kernel Ridge Regression (KRR), and Gaussian Process Regression (GPR) on data from skilled human performance of a variety of Brazilian… 

Figures from this paper

This work proposes a language model based on conditional random fields that integrates beat and onset likelihoods as observations and derives activations using deep neural networks and evaluates its performance on manually annotated data using a scheme adapted to this task.
Conditional Drums Generation using Compound Word Representations
A sequence-to-sequence architecture where a Bidirectional Long short-term memory (BiLSTM) Encoder receives information about the conditioning parameters (i.e., accompanying tracks and musical attributes), while a Transformer-based Decoder with relative global attention produces the generated drum sequences.
Modeling the tendency for music to induce movement in humans: first correlations with low-level audio descriptors across music genres.
Two descriptors corresponding to the density of events between beats and the salience of the beat, respectively, were strongly correlated with groove across domains, and systematic deviations from strict positions on the metrical grid, so-called microtiming, did not play any significant role.
A Cross-modal Heuristic for Periodic Pattern Analysis of Samba Music and Dance
A method is developed using a computational heuristics that provides an analysis framework to find periodic patterns connected with metre in dance and music, and it is shown that an interesting coupling can be observed between Dance and music at the metrical level.
In music, “groove” refers to the sense of rhythmic “feel” or swing. Groove, which was originally introduced to describe the taste of a bands rhythm section, has been expanded to non-rhythmic sections
PocketVAE: A Two-step Model for Groove Generation and Control
PocketVAE, a groove generation system that applies grooves to users’ rudimentary MIDI tracks, i.e, templates, takes a two-step approach that is analogous to a music creation process, and shows that the usage of a discrete encoding space improves the learning of the original data distribution.
Microtiming in “Samba de Roda” —Preliminary experiments with polyphonic audio
In this paper we report on preliminary experiments in the stu dy of microtiming features in Samba music. Focusing on polyphoni c audio music with almost constant tempo, we propose an algorithm to
Learning to Groove with Inverse Sequence Transformations
Though Seq2Seq models usually require painstakingly aligned corpora, it is shown that it is possible to adapt an approach from the Generative Adversarial Network (GAN) literature to sequences, creating large volumes of paired data by performing simple transformations and training generative models to plausibly invert these transformations.
Rhythm into style : studying assymmetrical grooves in Norwegian folk music
This thesis is a study of rhythm and timing in Norwegian traditional dance tunes called pols or springar in asymmetrical triple meter performed on the fiddle and Hardanger fiddle. This music is
Sonification of Samba dance using periodic pattern analysis
In this study we focus on the sonification of Samba dance, using a multi-modal analysis-by-synthesis approach. In the analysis we use periodic pattern analysis to decompose the Samba dance


Eigenrhythms: Drum pattern basis sets for classification and generation
This work took a collection of 100 drum beats from popular music tracks and normalized each pattern to form an ensemble of aligned drum patterns, using this low-dimension representation of the drum patterns as a space for classification and visualization, and discusses its application to generating continua of rhythms.
Ensemble Timing in Computer Music
The time map method allows the composer to specify a distortion of a time trajectory such that complex tempo variations combined with precise coincidence control can be obtained and is also a useful way to think about performance analysis.
Toward objective measurement and evaluation of jazz piano performance via MIDI-based groove quantize templates
The purpose of this study was to (a) objectively measure and analyze performance deviations from mechanical regularity for three jazz pianists via MIDI-based "groove quantize" procedures and (b)
Swing Ratios and Ensemble Timing in Jazz Performance: Evidence for a Common Rhythmic Pattern
The timing in jazz ensemble performances was investigated in order to approach the question of what makes the music "swing," suggesting a practical limit on tone duration that may be due to perceptual factors.
A case based approach to expressivity-aware tempo transformation
This work investigates the problem of how a performance played at a particular tempo can be rendered automatically at another tempo, while preserving naturally sounding expressivity, and presents a case-based reasoning system called TempoExpress for addressing this problem.
From Time to Time: The Representation of Timing and Tempo
  • H. Honing
  • Psychology
    Computer Music Journal
  • 2001
Timing plays an important role in the performance and appreciation of almost all types of music. It has been studied extensively in music perception and music performance research (see Palmer 1997
Machine learning
Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis.
“It Don’t Mean a Thing If It Ain’t Got That Swing” – Simulating Expressive Timing by Modulated Movements
A theoretical interpretation offered by this model is to view expressive timing as a result of rhythmic structure being “stretched” and “compressed” by actions of movements.
Embodied Mind, Situated Cognition, and Expressive Microtiming in African-American Music
The dual theories of embodied mind and situated cognition, in which physical/temporal embodiment and physical/social/cultural environment contribute crucially to the structure of mind, are brought to