Federico Simonetta

sound, machine learning, computing

fluid thoughts against categories

About

about.md

I am a post-doc researcher in the Laudare ERC project at the GSSI - Gran Sasso Science Institute. I am also a research collaborator at the University of Milan, in the LIM - Music Informatics Laboratory. Previously, I was a post-doc researcher at the Instituto Complutense de Ciencias Musicales - ICCMU in the Didone project , where I worked on the 18th century Italian Opera.

I have taken my Ph.D. in Computer Science at the University of Milan in the LIM - Music Informatics Laboratory , where I studied computational methods for music performance analysis. While I am continuously involved in researching new tools for enabling access to music/culture fruition/production, I want to contribute to the improvement of the society and of the world in which we live.

  • Main Research interests: music information processing, machine learning, audio processing, document understanding
  • Secondary Research interests: audio processing, medical acoustics, symbolic music analysis and generation, audio synthesis
  • CV: download it (old)

Publications

Here is the full list of publications .

The following are some notable publications that you may be interested in.

Screenshot of the Listening Test

See more

Federico Simonetta, S. Ntalampiras, F. Avanzini

Published in Multimedia Tools and Applications 2022

This study focuses on the perception of music performances when contextual factors, such as room acoustics and instrument, change. We propose to distinguish the concept of “performance” from the one of “interpretation”, which expresses the “artistic intention”. Towards assessing this distinction, we carried out an experimental evaluation where 91 subjects were invited to listen to various audio recordings created by resynthesizing MIDI data obtained through Automatic Music Transcription (AMT) systems and a sensorized acoustic piano. During the resynthesis, we simulated different contexts and asked listeners to evaluate how much the interpretation changes when the context changes. Results show that: (1) MIDI format alone is not able to completely grasp the artistic intention of a music performance; (2) usual objective evaluation measures based on MIDI data present low correlations with the average subjective evaluation. To bridge this gap, we propose a novel measure which is meaningfully correlated with the outcome of the tests. In addition, we investigate multimodal machine learning by providing a new score-informed AMT method and propose an approximation algorithm for the p-dispersion problem.


Acoustics-specific strategies improve velocity prediction

See more

Federico Simonetta, S. Ntalampiras, F. Avanzini

Published in MMSP 2022

Motivated by the state-of-art psychological research, we note that a piano performance transcribed with existing Automatic Music Transcription (AMT) methods cannot be successfully resynthesized without affecting the artistic content of the performance. This is due to 1) the different mappings between MIDI parameters used by different instruments, and 2) the fact that musicians adapt their way of playing to the surrounding acoustic environment. To face this issue, we propose a methodology to build acoustics-specific AMT systems that are able to model the adaptations that musicians apply to convey their interpretation. Specifically, we train models tailored for virtual instruments in a modular architecture that takes as input an audio recording and the relative aligned music score, and outputs the acoustics-specific velocities of each note. We test different model shapes and show that the proposed methodology generally outperforms the usual AMT pipeline which does not consider specificities of the instrument and of the acoustic environment. Interestingly, such a methodology is extensible in a straightforward way since only slight efforts are required to train models for the inference of other piano parameters, such as pedaling.


Benchmarks of feature extraction tools

See more

Federico Simonetta, Ana Llorens, Martín Serrano, Eduardo García-Portugués, Álvaro Torrente

Published in ISMIR 2023

This paper presents a comprehensive investigation of existing feature extraction tools for symbolic music and contrasts their performance to determine the feature set that best characterizes the musical style of a given music score. In this regard, we propose a novel feature extraction tool, named musif, and evaluate its efficacy on various repertoires and file formats, including MIDI, MusicXML, and **kern. Musif approximates existing tools such as jSymbolic and music21 in terms of computational efficiency while attempting to enhance the usability for custom feature development. The proposed tool also enhances classification accuracy when combined with other feature sets. We demonstrate the contribution of each feature set and the computational resources they require. Our findings indicate that the optimal tool for feature extraction is a combination of the best features from each tool rather than a single one. To facilitate future research in music information retrieval, we release the source code of the tool and benchmarks.


Full List Of Publications

All the code that I can publish is available in online. See my GitHub page for finding them. You can also check my Google Scholar profile .

PDFs of my papers are also available on my Zotero personal page whenever possible.


Teaching

Here I will put all the material I use for teaching. For now, it is just a list of useful links!



Theses

Here are all the theses that I have supervised so far. If you are interested in one of these topics, please contact me.

The Overleaf column is a reminder for myself, indicating if I have the source code of the LaTeX project on Overleaf.

See Table (not responsive)
Level Student Year Title Advisor Co-advisor Publication Link Overleaf
Bachelor Festorazzi Francesco 2019 Stima del parametro MIDI velocity da registrazioni audio polifoniche di pianoforte F. Avanzini me - - yes
Bachelor Talamona Francesco 2020 Una libreria C per il caricamento e la manipolazione di documenti IEEE 1599 L. A. Ludovico me - - yes
Bachelor Silipo Sebastiano 2021 Apprendimento one-shot per la classificazione dei generi musicali S. Ntalampiras me - - yes
Bachelor Nicolini Marco 2021 Audio-based Human Activity Classification using Transfer Learning S. Ntalampiras me conference doi yes
Master Cozzatti Michele 2021 Variational Autoencoders for Anomaly Detection in Respiratory Sounds S. Ntalampiras me conference arxiv yes
Master Poiré Alessandro Maria 2021 Ricerca Automatica di Features in Ambito di Acustica Medica S. Ntalampiras me conference arxiv yes
Bachelor Bellomo Carlotta 2022 Analysis of a convolutional model for melody-line identification S. Ntalampiras me - - yes
Bachelor Certo Francesca 2022 Analyzing emotion prediction system across different classes of sounds S. Ntalampiras me conference arxiv yes
Bachelor Giardina Alice 2022 Automatic classification of depression in speech S. Ntalampiras me - - yes
Master Pedretti Davide 2022 Sound classification in the presence of missing samples S. Ntalampiras me - - yes
Master Facchinetti Nicolas 2022 Adversarial Machine Learning in a Speech Emotion Recognition Scenario S. Ntalampiras me journal arxiv yes
Bachelor Peratello Emanuele 2023 Reti Neurali Bayesiane per l’Acustica Medica S. Ntalampiras me - - yes
Bachelor Demme Giulia 2024 Classificazione di suoni cardiaci mediante l’uso di Auto-Encoder Variazionale S. Ntalampiras me - - yes
Bachelor Tomaselli Leonardo 2024 Impatto della Window Length nella Classificazione di Suoni Respiratori Patologici S. Ntalampiras me - - yes
Master Cortis Paolo 2024 Adversarial Robustness Evaluation of Feature Learning Models and Universal Audio Representations S. Ntalampiras me - - yes
Master Palano Mattia 2024 Covid-19 Detection based on wavelet scatter transform S. Ntalampiras me - - yes
Bachelor Casalatina Oliviero ongoing Systematic Survey of Soundscape Synthesis Methods S. Ntalampiras me - - no
Bachelor Longhi Giorgio ongoing Interpretability of Medical Acoustic Models S. Ntalampiras me - - no
Master Mondal Rishav ongoing Optical Music Recognition for Cultural Heritage in the Ricordi Archive S. Ntalampiras me conference arxiv yes
Master Tabaraei Ali ongoing Domain generalization for multimodal speech-based depression detection S. Ntalampiras me - - no

Contacts

If you want, do not hesitate and write me an email!

My current local time is .

federico.simonetta [at] gssi.it