Prague Music Computing Group

We are a new research group that, as the name suggests, works on the computational processing and modeling of music in its various forms. Our strongest topics are Optical Music Recognition (OMR), and digital Gregorian chant scholarship. We are a part of the Institute of Formal and Applied Linguistics (UFAL).

The People

MgA. Jan Hajič jr., Ph.D. -- group lead and harpsichordist

Assoc. Prof. Pavel Pecina, Ph.D. -- in his capacity as PhD student supervisor

Mgr. Jiří Mayer -- Ph.D. student, Optical Music Recognition

Mgr. Adam Štefunko -- Ph.D. student, Computational models of basso continuo

Mgr. Vojtěch Lanz -- Ph.D. student, non-musical thesis but involved in chant scholarship as DACT project member -- melody segmentation with unsupervised Bayesian methods.

Bc. Anna Dvořáková -- Mgr. student, DACT project member. Bc. thesis on Analysing Gregorian chant repertoire traditions with clustering, community detection, and topic models (defended Sep. 2024).

Bc. Jan Borecký -- Mgr. student, music as navigational tool in open-world games for the visually impaired.

Bc. Filip Ruta -- Mgr. student, music notation learning game with a MIDI keyboard and generated content.

Bc. Kristýna Harvanová -- Mgr. student, Bc. thesis on synthesizing realistic images for Optical Music Recognition (defended Sep. 2024).

Bc. Patrik Backo -- Mgr. student, Bc. thesis on generating drum kit sounds for electronic music (defended Sep. 2024).

Reut Tal -- Bc. student, a game for collecting music emotion judgments in an immersive environment

Emre Rasimgil -- Bc. student, generating background music cheaply.

Vojtěch Dvořák -- Bc. student, image segmentation fast sheet music layout analysis

Šimon Libřický -- Bc. student, component for automatically incorporating arbitrary symbolic music processing tools into MuseScore

The Projects

OmniOMR (2023-2027): a project of the Ministry of Culture of the Czech Republic for developing an Optical Music Recognition system and deploying it at scale in the Moravian Library. PI: Jan Hajič.

DACT (2023-2030), Chant Analytics. Jan Hajič is a co-investigator of this 7-year Partnership Grant of the Social Sciences and Humanities Research Council of Canada (grant no. 895-2023-1002), leading the Chant Analytics team.

MASHCIMA (2023-2025): a project of the Grant Agency of the Charles University (GAUK) focusing on generating synthetic data for Optical Music Recognition. PI: Jiří Mayer.

Teaching

We start teaching the Computational Music Processing course (NPFL144) in 2024/2025.

If you are interested in a music-oriented thesis or individual software project, contact: hajicj@ufal.mff.cuni.cz

Note: We're still looking for students for the international SCORIA student project with the Institute of Computational Perception in Linz.

The Events

November 28th, 2024: Jan Hajič jr. will represent PMCG at the Open Doors Day of the Faculty of Mathematics and Physics.

February 13th-14th, 2025: Workshop on Romani Chords and digital tools for recording performance: accordions, guitars, ... Organized by Petr Nuska at the Ethnological Institute of the Czech Academy of Sciences.

April 25th, 2025: Prague Music Computing Day no. 2 (tentative)

(Past)

September 5th, 2024: Three of our students defend their Bc. theses: Anna Dvořáková (mapping the repertoire of Gregorian Chant), Patrik Backo (Neural drum one-shot synthesis) and Emre Rasimgil (generating low-cost elevator music).

September 19th, 2024: We go to Vienna! Jan Hajič and Adam Štefunko are presenting PMCG work for the

June 6th, 2024: DACT Chant Analytics Workshop

April 19th, 2024: 1st PMCG Workshop