Next Article in Journal / Special Issue
An Interpretable Deep Learning Model for Automatic Sound Classification
Previous Article in Journal
UFMC Waveform and Multiple-Access Techniques for 5G RadCom
Previous Article in Special Issue
Investigating the Effects of Training Set Synthesis for Audio Segmentation of Radio Broadcast
Article

Informing Piano Multi-Pitch Estimation with Inferred Local Polyphony Based on Convolutional Neural Networks

Fraunhofer IDMT, Semantic Music Technology Group (SMT), 98693 Ilmenau, Germany
*
Author to whom correspondence should be addressed.
Academic Editors: Alexander Lerch and Peter Knees
Electronics 2021, 10(7), 851; https://doi.org/10.3390/electronics10070851
Received: 28 February 2021 / Revised: 26 March 2021 / Accepted: 27 March 2021 / Published: 2 April 2021
(This article belongs to the Special Issue Machine Learning Applied to Music/Audio Signal Processing)
In this work, we propose considering the information from a polyphony for multi-pitch estimation (MPE) in piano music recordings. To that aim, we propose a method for local polyphony estimation (LPE), which is based on convolutional neural networks (CNNs) trained in a supervised fashion to explicitly predict the degree of polyphony. We investigate two feature representations as inputs to our method, in particular, the Constant-Q Transform (CQT) and its recent extension Folded-CQT (F-CQT). To evaluate the performance of our method, we conduct a series of experiments on real and synthetic piano recordings based on the MIDI Aligned Piano Sounds (MAPS) and the Saarland Music Data (SMD) datasets. We compare our approaches with a state-of-the art piano transcription method by informing said method with the LPE knowledge in a postprocessing stage. The experimental results suggest that using explicit LPE information can refine MPE predictions. Furthermore, it is shown that, on average, the CQT representation is preferred over F-CQT for LPE. View Full-Text
Keywords: polyphony estimation; multi-pitch estimation; convolutional neural networks; music information retrieval polyphony estimation; multi-pitch estimation; convolutional neural networks; music information retrieval
Show Figures

Figure 1

  • Externally hosted supplementary file 1
    Doi: 10.5281/zenodo.4637908
    Link: https://doi.org/10.5281/zenodo.4637908
    Description: We have now published the dataset SMD-synth, which is described in Section 5.3 of the paper. This has also been noted in the Data Availability Statement of the submitted (revised) paper, and a respective reference has been placed under the Reference section. The dataset is accessible under the given link.
MDPI and ACS Style

Taenzer, M.; Mimilakis, S.I.; Abeßer, J. Informing Piano Multi-Pitch Estimation with Inferred Local Polyphony Based on Convolutional Neural Networks. Electronics 2021, 10, 851. https://doi.org/10.3390/electronics10070851

AMA Style

Taenzer M, Mimilakis SI, Abeßer J. Informing Piano Multi-Pitch Estimation with Inferred Local Polyphony Based on Convolutional Neural Networks. Electronics. 2021; 10(7):851. https://doi.org/10.3390/electronics10070851

Chicago/Turabian Style

Taenzer, Michael, Stylianos I. Mimilakis, and Jakob Abeßer. 2021. "Informing Piano Multi-Pitch Estimation with Inferred Local Polyphony Based on Convolutional Neural Networks" Electronics 10, no. 7: 851. https://doi.org/10.3390/electronics10070851

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop