Music education technologies developed by MTG
Music Critic platform, developed by the Music Technology Group (MTG), Universitat Pompeu Fabra, serves as a showcase for music education demos. This platform highlights the innovative work of the MTG in the field of music education, featuring various demos and resources aimed to promote the research on helping individuals enhance their musical knowledge and skills.
DEMOS
Difficulty Estimation from PDF
Picking the right piano piece in a music library? Our model handles sheet music images or PDFs directly to predict its difficulty. It is trained on the vast IMSLP piano collection and fine-tuned in smaller difficulty datasets.
DEMODifficulty Estimation from MusicXML
This demo will predict the difficulty of a piano piece based on various input representations from score information to pre-trained ML models for generative piano performance inspired by the Nicholas Cook definition of performance.
DEMOPiano Fingering Estimation from MusicXML
This demo uses neural networks to predict the piano fingering for music scores in MusicXML format. Knowing the multiple possible fingerings of a piece is fundamental to understanding its technique.
DEMOOnset Detection
This real-time onset detection demo operates directly in your web browser, identifying the beginning of musical notes or events. It's perfect for musicians, producers, and researchers who want to analyze and visualize the timing of musical events without specialized software or equipment.
DEMOMikrokosmos-difficulty feedback
In this paper we present score difficulty classification as a way to improve music education technology. Using a novel dataset, Mikrokosmos-difficulty, two ML models are compared. The models generate interpretable difficulty feedback for piano pieces, aiding in personalized curriculum development and score retrieval.
DEMO
Real-time Pitch Detection
Discover our real-time pitch detection demo that operates directly in your web browser. This interactive demo allows you to input sound from your microphone or an audio file, and instantly observe detected pitches as you play or sing. The innovative technology behind this demo showcases its potential for various musical applications, including online music education, transcription, and creative sound manipulation. Try it out now!
DEMOChords analysis
Real-time chord analysis demo: Analyze the chords of the music you're playing in real-time directly in your web browser with this interactive demo. Simply play a song on your instrument or device, and the demo will detect the chords and display them in real-time. Ideal for musicians looking to improve their chord recognition skills or for music educators to demonstrate chord analysis techniques.
DEMO