Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Фундаментальные иссл...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
addClaim

Нейросетевое распознавание музыкальных инструментов с использованием мел‑частотных кепстральных коэффициентов

Нейросетевое распознавание музыкальных инструментов с использованием мел‑частотных кепстральных коэффициентов

Abstract

В данной статье рассматривается распознавание образов с применением нейронных сетей на примере распознавания музыкальных инструментов. Проблема распознавания звука музыкальных инструментов рассматривается во многих работах, но большинство из них используют большое количество признаков. Целью нашей работы является минимизация количества признаков при условии получения высокой точности распознавания. В качестве признаков используются мел-частотные кепстральные коэффициенты с применением метода главных компонент. Для классификации признаков используется нейронная сеть прямого распространения. Для обучения сети были рассмотрены два подхода: метод обратного распространения ошибки и нейроэволюционный алгоритм Enforced Subpopulations (ESP). Метод обратного распространения ошибки показал лучшие результаты. Данная работа посвящена распознаванию отдельно звучащих нот музыкальных инструментов, однако предложенный алгоритм показал хорошие результаты и при анализе сложных сигналов.

In this paper the task of automatic musical instrument recognition is considered. A lot of papers have been published on this topic, but most of them use a great number of features for recognition of musical instrument sound. This work is devoted to minimization the number of features together with getting of high recognition accuracy. The Mel-Frequency Cepstral Coefficients are considered as the main set of features. Additionally, the number of MFCC has been optimized using Principle Component Analysis. A feedforward neural network was used as a classifier. Two approaches were considered for classifier training: backpropagation method and Enforced Subpopulations neuroevolution approach. The backpropagation method was determined as the best one. Basically, this work is devoted to recognition of isolated notes, but also the developed algorithm was applied to a complex signal and as a result solo parties of musical instruments were successfully identified.

Keywords

РАСПОЗНАВАНИЕ ОБРАЗОВ, МУЗЫКАЛЬНЫЕ ИНСТРУМЕНТЫ, МЕЛ-ЧАСТОТНЫЕ КЕПСТРАЛЬНЫЕ КОЭФФИЦИЕНТЫ, МЕТОД ГЛАВНЫХ КОМПОНЕНТ, НЕЙРОЭВОЛЮЦИЯ

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
gold