Research
Research
The starting point of my research is almost always a musical question — a problem that arises from compositional practice rather than from technology. That question is translated into a mathematical formalization, which suggests an implementation — an algorithm, increasingly one built on machine learning — that turns the formal model into a tool I can compose with and test against real scores and live performance. This is not a one-way pipeline but a feedback loop: research clarifies the art, the art redirects the research, and the two advance together rather than in sequence.
Computer-assisted orchestrationTarget-based orchestration and the Orchidea framework: reconstructing a sound with a real orchestra by searching a space of instrumental combinations.
Timbre and representationMathematical models of musical signals: scattering transforms, symbolic representations (sound-types), timbre spaces.
Machine learning for soundText-to-sound mapping, sound hybridisation, network bending of diffusion models, style transfer: controllable, interpretable tools for artists rather than opaque generative black boxes.
Augmented instrumentsEmbedded devices capable of reshaping the acoustic identity of an instrument as a function of the performer's gesture, and of linking several instruments into a collective hyper-instrument (Les espaces physiques, I am in blood, the augmented vibraphone).
Selected publications
2026
- Computer-assisted orchestration: on the genesis of ideas at IRCAM (2003–2006) (with Yan Maresz). Oxford Handbook of Orchestration Studies, Oxford University Press.
- InstructFX2FX: a multi-turn text-to-effect system for sequential audio effect refinement (Song-Ze Yu, Milan Liessens Dujardin, Yuxuan Cai, Wantong Zhang, Brian Cruz, Jeremy Wagner, C.-E. Cella). DAFx26, MIT, Boston.
- Evaluating tokenization strategies for expressive classical piano performance generation (Qingyang Lyu, Brian Cruz, Jeremy Wagner, C.-E. Cella). DAFx26, MIT, Boston.
- Segmentation-free sound hybridization for creative outcomes (Pasquale Mainolfi, C.-E. Cella). EvoMUSART 2026, Toulouse.
2025
- Mimésis, catharsis, et l'avenir de l'IA dans la création musicale. In L'Art au temps de l'IA, ed. J.-L. Giavitto and P. Saint-Germier, Éditions du Centre Pompidou (in French) — book.
- Generating music-reactive videos by applying network bending to Stable Diffusion (Luke Dzwonczyk, C.-E. Cella, David Ban). Journal of the Audio Engineering Society, 73(6).
- An overview on CNMAT technologies and future directions (with Edmund Campion, Jeremy Wagner, Luke Dzwonczyk, Jon Kulpa). ICMC 2025, Boston.
- A dynamically controlled feedback system: Waking cloud (Nicola Casetta, C.-E. Cella). ICMC 2025, Boston.
- Unsupervised text-to-sound mapping via embedding space alignment (Luke Dzwonczyk, C.-E. Cella). DAFx25, Ancona.
2024
- Network bending of diffusion models for audio-visual generation (with Luke Dzwonczyk and David Ban). DAFx24, Guildford.
- Audio visualization via delay embedding and subspace learning (with Alois Cerbu). DAFx24, Guildford.
- Reprogrammable effects pedals on the Daisy Seed platform (with Alois Cerbu). ICMC 2024, Seoul.
2022
- Orchidea: a comprehensive framework for target-based computer-assisted dynamic orchestration. Journal of New Music Research.
- Neural models for target-based computer-assisted musical orchestration: a preliminary study (with Luke Dzwonczyk, Alejandro Saldarriaga-Fuertes, Hongfu Liu, Hélène-Camille Crayencour). Journal of Creative Music Systems.
- A three-dimensional timbre model via Peano curves (with Daniele Ghisi). Journal of Mathematics and Music.
- Color and timbre gestures: an approach with bicategories and bigroupoids (with Maria Mannone, Giovanni Santini, Esther Adedoyin). MDPI Mathematics, 10(4).
- Synthesis by layering: learning a variational space of drum sounds (with Elliott Waissbluth and Jon Gillick). AI Music Creativity 2022.
- Vivace: web application for real-time feedback on piano performance (with Jeremy Lee and Hélène-Camille Crayencour). AI Music Creativity 2022.
- Source separation methods for computer-assisted orchestration (with Luke Dzwonczyk, Léo Chédin, Alejandro Saldarriaga-Fuertes, Max Sherr, Hélène-Camille Crayencour). AI Music Creativity 2022.
2021
- Dynamic computer-aided orchestration in practice with Orchidea (with Daniele Ghisi, Yan Maresz, Alessandro Petrolati, Alexandre Teiller, Philippe Esling). Computer Music Journal.
- A framework for modifying orchestral qualities in computer-aided orchestration (with Daniele Ghisi). Computer Music Journal.
- Drumroll please: modeling multi-scale rhythmic gestures with flexible grids (with Jon Gillick, Joshua Yang, David Bamman). Transactions of ISMIR.
- Imitative computer-aided musical orchestration with biologically inspired algorithms (with Marcelo Caetano). Handbook of Artificial Intelligence in Music, Springer.
2020
- A study on neural models for target-based computer-assisted musical orchestration (with Luke Dzwonczyk, Alejandro Saldarriaga-Fuertes, Hongfu Liu, Hélène-Camille Crayencour). AI Music Creativity 2020, Stockholm.
- Music information retrieval and contemporary classical music: a successful failure. Transactions of ISMIR, 3(1).
- OrchideaSOL: a dataset of extended instrumental techniques for computer-aided orchestration (with Daniele Ghisi, Vincent Lostanlen, Fabien Lévy, Joshua Fineberg, Yan Maresz). ICMC 2020, Santiago.
- QuBits, a system for interactive sonic virtual reality (with Jon Kulpa and Edmund Campion). ICMC 2020, Santiago.
- Kymatio: scattering transforms in Python (with M. Andreux, T. Angles, G. Exarchakis, R. Leonarduzzi, G. Rochette, L. Thiry, J. Zarka, S. Mallat, J. Andén, E. Belilovsky, J. Bruna, V. Lostanlen, M. Chaudhary, M. J. Hirn, E. Oyallon, S. Zhang, M. Eickenberg). Journal of Machine Learning Research, 21(60).
2019
- Estimating unobserved audio features for target-based orchestration (with Jon Gillick and David Bamman). ISMIR 2019, Delft.
- Learning, probability and logic: towards a unified approach for content-based music information retrieval (with Hélène-Camille Crayencour). Frontiers in Digital Humanities.
2018
- Open-source modular toolbox for computer-aided orchestration (with Philippe Esling). Timbre 2018, Montreal.
- Deep learning for timbre modification and transfer: an evaluation study (with Leonardo Gabrielli, Fabio Vesperini, Diego Droghini, Emanuele Principi, Stefano Squartini). AES 144th Convention, Milan.
Earlier
- Machine listening intelligence. International Workshop on Deep Learning for Music, Anchorage, 2017.
- Deep convolutional networks on the pitch spiral for musical instrument recognition (with Vincent Lostanlen). ISMIR 2016, New York.
- Vuza: a functional language for creative applications. ICMC 2014, Athens — project page.
- Advanced sound hybridizations by means of the theory of sound-types (with Juan José Burred). ICMC 2013, Perth — audio examples.
- Sound-types: a new framework for symbolic sound analysis and synthesis. ICMC 2011, Huddersfield.
- Nuovi approcci alla struttura armonica: caleidocicli e mosaici tricordali. In Luigi Verdi, Caleidocicli musicali, 2nd ed., Rugginenti, Milan, 2010.
- Harmonic components extraction in recorded piano tones. AES 128th Convention, London, 2010.
- Towards a symbolic approach to sound analysis. Mathematics and Computation in Music, Yale University, 2009.
- Using the SDIF Sound Description Interchange Format for audio features (with J. J. Burred, G. Peeters, A. Röbel, D. Schwarz). ISMIR 2008, Philadelphia.
- Recensione di Caleidocicli musicali di Luigi Verdi. Musica Theorica Spectrum, Curci, 2006.
- Il compositore cieco. Rivista umbra di musicologia, 50, 2006.
- Il semplice sistema. Rivista umbra di musicologia, 49, 2005.
- Sulla struttura logica della musica. Rivista umbra di musicologia, 48, 2005 — presented at IRCAM's MaMuX seminar, 2006; extended version.
- Sintesi per stati e visualizzazione del processo compositivo (with Federico Paolinelli). La Terra Fertile, L'Aquila, 2000.
Theses and other materials
- On symbolic representations of music. PhD dissertation, University of Bologna, 2011 — defence slides; a short description of the sound-types transform; presentation at Queen Mary University of London.
- Sulla struttura logica della musica. Master's thesis in philosophy (mathematical logic), University of Urbino, 2004.
Selected lectures
- 2025 — Collège de France, opening symposium Formes de l'intelligence: De la mimesis à la katharsis
- 2021 — McGill University, CIRMMT Distinguished Lecture: On computational creativity
- 2019 — CCRMA, Stanford University; Columbia University; Conservatoire de Paris; HEM Genève
- 2017 — Institute of Mathematics, National Academy of Sciences of Ukraine; Schloss Dagstuhl
- 2015 — École normale supérieure, Paris; MaMuX, IRCAM
- 2011 — Queen Mary University of London; IRMA, Strasbourg
Service
Reviewer and programme-committee member for the Computer Music Journal, EURASIP JASP, ISMIR, MCM, CIM, IEEE TETCI and the IRCAM musical research residency; member of the EURASIP Special Area Team on Acoustic, Speech and Music Signal Processing; organiser of the Re.M.I.X round-tables on music research and of the CNMAT OpenLab.