Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

This Research product is the result of merged Research products in OpenAIRE.

You have already added 0 works in your ORCID record related to the merged Research product.

Исследование алгоритмов по распознаванию высоты звуков в Ð¼ÑƒÐ·Ñ‹ÐºÐ°Ð»ÑŒÐ½Ñ‹Ñ Ñ„Ñ€Ð°Ð³Ð¼ÐµÐ½Ñ‚Ð°Ñ

выпускная квалификационная работа бакалавра

Исследование алгоритмов по распознаванию высоты звуков в Ð¼ÑƒÐ·Ñ‹ÐºÐ°Ð»ÑŒÐ½Ñ‹Ñ Ñ„Ñ€Ð°Ð³Ð¼ÐµÐ½Ñ‚Ð°Ñ

Abstract

В данной работе приводится математическое описание четырёх наиболее распространённых алгоритмов по распознаванию нот в одноголосных мелодиях, осуществляется их программная реализация, производится анализ их точности и эффективности на распознавании аудиозаписей, в которых помимо мелодии содержатся “шумовые” отзвуки и обертоны, а также специфические приёмы игры, характерные для определённых музыкальных инструментов. В процессе интерпретации результатов распознавания выявляются несовершенства алгоритмов при работе с такими аудиозаписями. Данные несовершенства объясняются с точки зрения специфики алгоритмов и методов цифровой обработки сигналов, использованных в их основе. В завершение анализа алгоритмов по распознаванию одноголосных мелодий приводится их сравнительная характеристика по всем рассмотренным аспектам в соответствии с использованными для работы аудиозаписями. Далее в работе разрабатывается нейроалгоритм для смежной задачи: распознавания аккордов в музыкальных фрагментах. После описания реализации приводится оценка качества и универсальности работы алгоритма.

This paper gives mathematical descriptions of four mostly widespread single-pitch estimation algorithms, and then describes their software implementation. After that, the accuracy and effectiveness of these algorithms are measured based on the results captured after processing audio-recordings with different «noisy» sounds, overtones, echo effects, timbre features and music techniques of various musical instruments. Obviously, all these aspects affect the estimation accuracy. The goal of the part dedicated to single pitch estimation algorithms is to analyze the accuracy loss and relate the loss with the steps implemented in the algorithms. The end of this part of work provides the comparative characteristics based on the aspects chosen in accordance with the features of audio-recordings being processed. The remainder of the work is dedicated to the implementation of the neuro-algorithm for the related purpose: multi-pitch estimation. The implementation description is followed by the quality and operational universality evaluation of the model.

Keywords

сравнительный анализ, python, keras, цифровая обработка сигналов, comparative analysis, neural network, оценка качества, digital signal processing, распознавание частоты звука, pitch estimation, нейронная сеть, quality evaluation

  • BIP!
    Impact byBIP!
    citations
    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
citations
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
Upload OA version
Are you the author? Do you have the OA version of this publication?