Research lines

  • Audio processing
  • Machine learning
  • Sound synthesis
  • Computer music
  • Music performance
  • Computational musicology
  • Computational perception

 

Projects

  • AI & Music - Chair on AI and Music (NextGenerationEU, 2023-2026). Researcher. Harnessing the potential of AI by developing practical applications and training professionals capable of leading the transformation and renewal of the music industry.
  • Candombe·an·do - Candombe analysis and documentation (CSIC-UdelaR, 2021-2022). Principal investigator. The purpose of this project is to document and analyse various aspects of candombe drumming through computational techniques.
  • MOSMO - Montevideo Sound Monitoring (UdelaR-IdeM, 2020-2021). Scientific coordinator and researcher. The project aim is to monitor noise pollution and detect annoying sounds in Montevideo city by using a network of sensors and signal processing and machine learning.
  • StaRel - Statistical Relational Learning for Music (STIC-AmSud, 2018-2019). Scientific coordinator and researcher. The project aims to gain insight into the understanding and modeling of the rhythmic/metrical structure in audio recordings of expressive music performances.
  • IEMP - Interpersonal Entrainment in Music Performance (AHRC, 2016-2018). Visiting fellow and collaborator. The project aims to develop a better understanding of how groups of people coordinate behaviour within a musical context, considering a broad range of musical cultures.

 

Code

  • carat - Computer-aided Rhythm Analysis Toolbox (python). A toolbox for rhythmic pattern analysis from audio recordings, including some ready-to-use software applications.
  • DCASE-models - Environmental sound analysis (python). An open-source library for prototyping environmental sound analysis systems, with an emphasis on deep-learning models.
  • FChT - Fan Chirp Transform and F0gram (Matlab, C). Improved time-frequency representation of polyphonic music and fundamental-frequency salience visualisation.
  • IIR-CQT - IIR Constant-Q Transform (Matlab). Constant-Q transform implemented by IIR filtering of the spectrum, as a good compromise between flexibility and low computational cost.
  • WIST - Wind Instruments Synthesis Toolbox (Matlab). Wind instruments additive synthesis for generation of music audio signals with labeled partials.

 

Data

  • IEMP Uruguayan Candombe - A collection of audiovisual recordings of Uruguayan Candombe drum ensemble performances, with detailed annotations (beat/downbeats, onsets).
  • Uruguayan Candombe Beat and Downbeat - A dataset of audio recordings of Uruguayan Candombe drum ensemble performances with annotated beats and downbeats.
  • Urban Sound & Sight (Urbansas) - A dataset for the development and evaluation of machine listening systems for audiovisual spatial urban understanding.
  • MAVD-traffic dataset - A dataset for sound event detection in urban environments that focuses on traffic noise, with audio recordings and synchronized video files.