Score informed audio source separation using constrained nonnegative matrix factorization and score synthesis
- 1 May 2013
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15206149,p. 888-891
Abstract
In this paper we present a new method for musical audio source separation, using the information from the musical score to supervise the decomposition process. An original framework using nonnegative matrix factorization (NMF) is presented, where the components are initially learnt on synthetic signals with temporal and harmonic constraints. A new dataset of multitrack recordings with manually aligned MIDI scores is created (TRIOS), and we compare our separation results with other methods from the literature using the BSS EVAL and PEASS evaluation toolboxes. The results show a general improvement of the BSS EVAL metrics for the various instrumental configurations used.Keywords
This publication has 8 references indexed in Scilit:
- Using score-informed constraints for NMF-based source separation2012
- Algorithms for Nonnegative Matrix Factorization with the β-DivergenceNeural Computation, 2011
- Soundprism: An Online System for Score-Informed Source Separation of Music AudioIEEE Journal of Selected Topics in Signal Processing, 2011
- Score informed audio source separation using a parametric model of non-negative spectrogram2011
- Subjective and Objective Quality Assessment of Audio Source SeparationIEEE Transactions on Audio, Speech, and Language Processing, 2011
- Monaural Sound Source Separation by Nonnegative Matrix Factorization With Temporal Continuity and Sparseness CriteriaIEEE Transactions on Audio, Speech, and Language Processing, 2007
- Performance measurement in blind audio source separationIEEE Transactions on Audio, Speech, and Language Processing, 2006
- Non-negative matrix factorization for polyphonic music transcription2004