Realtime Transcription
From IMTR
NEWS
- Book chapter on Real-time detection of overlapping sound events with non-negative matrix factorization now in press, draft available.
- Our real-time system for polyphonic music transcription performed at the top during MIREX 2010 where it was in competition with off-line algorithms. See our ISMIR 2010 paper.
This page demonstrates the real-time transcription technology developed at Ircam for polyphonic audio. Transcription is in short the mapping of audio signals to symbolic values, such as notes, commonly used by musicians.
The technology presented here is based on non-negative decomposition techniques. See Further Readings.
Demo
Following is the demo of the first prototype presented in 2006. This prototype was based on sparse non-negative decomposition techniques. Our recent researches have focused on non-negative decomposition with the beta-divergence to improve transcription results. These researches have been presented during summer 2010. New demos and code using our most recent techniques will come soon. Stay tuned...
Contributors
- Research and Development: Arshia Cont, Arnaud Dessein
Further Readings
- Arnaud Dessein, Arshia Cont, and Guillaume Lemaitre. Real-time detection of overlapping sound events with non-negative matrix factorization. In Frank Nielsen and Rajendra Bhatia, editors, Matrix Information Geometry, chapter 14, pages 341–371. Springer, Berlin/Heidelberg, Germany, 2013. (draft) (bibtex)
- Arnaud Dessein, Arshia Cont, and Guillaume Lemaitre. Real-time polyphonic music transcription with non-negative matrix factorization and beta-divergence. In 11th International Society for Music Information Retrieval Conference (ISMIR), pages 489–494, Utrecht, Netherlands, August 2010. (paper) (bibtex) (poster)
- Arnaud Dessein, Arshia Cont, and Guillaume Lemaitre. Real-time polyphonic music transcription with non-negative matrix factorization and beta-divergence. In 6th Music Information Retrieval Evaluation eXchange (MIREX), Utrecht, Netherlands, August 2010. (abstract) (bibtex) (web)
- Arnaud Dessein. Incremental multi-source recognition with non-negative matrix factorization. Master's thesis, Université Pierre et Marie Curie, Paris, France, June 2009. (report) (bibtex) (slides) (web)