🎙️ Audio Explorer | 🤖 Machine Learning Enthusiast

I am Matteo Lionello, a passionate researcher and developer specializing in the intersection of machine learning, deep learning, audio technologies and neuroimaging. Currently, I am working in Cognitive, Computational, and Social Neuroscience at IMT Lucca, I have a diverse background that spans urban soundscapes, music technology, and computer science.

🔗 LinkedIn | 🌐 Personal Website

• Ph.D. Student, IMT Lucca

  • Cognitive, Computational, and Social Neuroscience

• MPhil, The Bartlett Institute, UCL

  • Thesis: A new methodology for modeling urban soundscapes

• MSc in Sound and Music Computing, Aalborg University

  • Thesis: Deep Learning for Sounds Representation and Generation

• BSc in Information Engineering, University of Padova

  • Thesis: Interactive Soundscapes

• Programming Languages:

 1. rwx Python       ██████████████░░
 2. rwx Matlab       ██████████████░░
 3. rwx Unix         ███████████░░░░░
 4. rwx Java         ██████████░░░░░░
 5. r-x C++          ████░░░░░░░░░░░░

• Libraries:

 1. rwx Tensorflow          █████████████░░░
 2. rwx Pytorch             █████████████░░░
 3. rwx Keras               ██████████░░░░░░
 4. r-x Pytorch Lightning   ████░░░░░░░░░░░░

Other skills (unranked):

• Protocols: OpenSoundsControl (OSC) and TCP/IP
• Databases: Basic knowledge of SQL and MySQL
• Tools: AFNI, Git, PureData, Android, HTML, CSS, and tiny bit of PhP

Audio quality enhancement and real-time audio denoising, developed as part of a project for a private company.

• Developed a Bayesian-NN for individual soundscape assessment predictions.
• Psychometric revisitation of the ISO protocol for soundscape measuring.

Collaborated with the Music Technology Group at Universitat Pompeu Fabra, Barcelona, on a machine learning approach to violin vibrato modeling. Developed a GUI on Android for real-time pitch detection.

Implemented LSTM modeling for melody and rhythmic structure extraction and generation during my time at Aalborg University.

Developed an end-to-end dilated variational autoencoder for sound morphing using TensorFlow.

• Created an interactive soundscape-augmented space at the University of Padua, inspired by Truax's work. Used granular synthesis to reproduce a collage of sounds referring to a real soundscape.

• Participated in a competitive workshop at Aalborg University, developing solutions for a new lighting-sound system at the AAU University bridge.

Drop me a message on LinkedIn or swing by my Website for all the details on my audio escapades!