本文關(guān)鍵詞：基于神經(jīng)網(wǎng)絡(luò)的聲樂演奏評(píng)價(jià)系統(tǒng)研究及其在鋼琴教學(xué)中的應(yīng)用　出處：《貴州大學(xué)》2008年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 神經(jīng)網(wǎng)絡(luò) 鋼琴教育 音樂評(píng)價(jià) MIDI 電腦音樂 音樂特征

【摘要】： 隨著近年來鋼琴教學(xué)的興起,大量的人員加入到了學(xué)習(xí)鋼彈奏的隊(duì)伍。但昂貴的鋼琴教學(xué)費(fèi)用以及它特有的教師和學(xué)生一對(duì)一的教學(xué)模式造成了鋼琴教育資源非常的緊缺,學(xué)習(xí)鋼琴演奏成為了一項(xiàng)奢侈的活動(dòng)。于是采用電腦多媒體軟件進(jìn)行鋼琴教學(xué)就成為了緩解該矛盾的一種可行的方式。本文討論了鋼琴教學(xué)軟件實(shí)現(xiàn)方法,對(duì)電腦鋼琴教學(xué)中的難點(diǎn)(即電腦教學(xué)是單向的知識(shí)傳授而缺少交互的環(huán)節(jié))提出了采用神經(jīng)網(wǎng)絡(luò)模型對(duì)鋼琴演奏進(jìn)行評(píng)價(jià)的方法,并用來模擬教師指導(dǎo)學(xué)生進(jìn)行彈奏練習(xí)。 對(duì)于一首音樂的演奏,影響其效果的因素是多種多樣的,對(duì)其評(píng)價(jià)的指標(biāo)也有很多種,如節(jié)奏感、表現(xiàn)力、樂感、風(fēng)格的把握等。采用電腦來模擬這個(gè)評(píng)價(jià)過程,實(shí)質(zhì)上就是要找出影響音樂演奏效果的因素和評(píng)價(jià)指標(biāo)之間的數(shù)理關(guān)系。神經(jīng)網(wǎng)絡(luò)是人工智能里通過模擬人類大腦思維方式提出來的一種數(shù)學(xué)模型,它具有對(duì)數(shù)據(jù)分布要求不嚴(yán)格、非線性的數(shù)據(jù)處理方法、強(qiáng)魯棒性和動(dòng)態(tài)性等優(yōu)點(diǎn),非常適合作為評(píng)價(jià)系統(tǒng)的數(shù)學(xué)模型。另外神經(jīng)網(wǎng)絡(luò)也有很強(qiáng)的理論基礎(chǔ),其在各行業(yè)中的應(yīng)用也發(fā)展得基本成熟,本文嘗試著把神經(jīng)網(wǎng)絡(luò)數(shù)學(xué)模型引入到鋼琴演奏的評(píng)價(jià)系統(tǒng)中來。 本論文所完成的主要工作: 1、論文首先說明了神經(jīng)網(wǎng)絡(luò)的原理,然后根據(jù)音樂理論找出了影響演奏效果的因素,包括音符、節(jié)奏、節(jié)拍、旋律、調(diào)性等。然后對(duì)每一種因素進(jìn)行量化,并作為神經(jīng)網(wǎng)絡(luò)的輸入?yún)?shù)。 2、根據(jù)音樂演奏的評(píng)價(jià)指標(biāo)設(shè)計(jì)神經(jīng)網(wǎng)絡(luò)模型,在鋼琴教師和學(xué)生的幫助下獲得了鋼琴演奏樣本數(shù)據(jù),完成了神經(jīng)網(wǎng)絡(luò)的訓(xùn)練。 3、介紹了現(xiàn)階段鋼琴教學(xué)軟件的基本情況,設(shè)計(jì)了一個(gè)鋼琴教學(xué)軟件的框架。然后實(shí)現(xiàn)了《歡樂頌》鋼琴彈奏練習(xí)的功能,并采用神經(jīng)網(wǎng)絡(luò)評(píng)價(jià)模塊對(duì)教師和學(xué)生的彈奏進(jìn)行了評(píng)價(jià),檢驗(yàn)?zāi)Ｐ偷男阅堋?鋼琴學(xué)習(xí)的初期功能完備的MIDI鍵盤和真實(shí)的鋼琴相比對(duì)于演奏者的發(fā)揮并無很大的影響,因而彈奏系統(tǒng)采用了MIDI鍵盤作為獲取演奏特征的工具。彈奏軟件采用了VC++編寫實(shí)現(xiàn),運(yùn)行于Win32平臺(tái)上面。在六盤水“音韻琴行”邀請(qǐng)了一位鋼琴教師和兩位學(xué)生進(jìn)行了彈奏試驗(yàn)。在比對(duì)了他們的不同的彈奏結(jié)果后得出,系統(tǒng)設(shè)計(jì)的各項(xiàng)指標(biāo)均達(dá)到了要求。
[Abstract]:With the rise of piano teaching in recent years. A large number of people have joined the team to learn how to play steel. But the high cost of piano teaching and its unique one-to-one teaching mode between teachers and students cause the shortage of piano education resources. Learning to play piano has become an extravagant activity. Therefore, the use of computer multimedia software for piano teaching has become a feasible way to alleviate this contradiction. This paper discusses the implementation method of piano teaching software. This paper puts forward a method of evaluating piano playing by using neural network model, which is difficult in computer piano teaching (that is, computer teaching is one-way knowledge imparting but lack of interactive link). And used to simulate the teacher to guide students to play. For the performance of a music, the factors that affect its effect are various, and there are many kinds of evaluation indicators, such as rhythm, expressiveness, musical sense. Use a computer to simulate the evaluation process. In essence, it is necessary to find out the mathematical relationship between the factors that affect the effect of music playing and the evaluation index. Neural network is a mathematical model proposed by simulating the way of thinking of human brain in artificial intelligence. It has the advantages of not strict data distribution, nonlinear data processing, strong robustness and dynamic, etc. It is very suitable for the mathematical model of evaluation system. In addition, the neural network also has a strong theoretical basis. Its application in various industries has also developed maturely. This paper attempts to introduce the neural network mathematical model into the evaluation system of piano performance. The main work of this thesis is as follows: 1. Firstly, the paper explains the principle of neural network, and then finds out the factors that affect the performance according to the music theory, including notes, rhythm, rhythm, melody, tonality and so on. Then it quantifies each factor. And as the input parameters of the neural network. 2. According to the evaluation index of music performance, the neural network model is designed. With the help of piano teachers and students, the sample data of piano performance are obtained, and the training of neural network is completed. 3. This paper introduces the basic situation of piano teaching software at present, designs a framework of piano teaching software, and then realizes the function of piano playing practice. The neural network evaluation module is used to evaluate the performance of the model. In the early stage of piano learning, the MIDI keyboard has no great influence on the player's performance compared with the real piano. Therefore, the MIDI keyboard is used as the tool to obtain the performance features in the playing system, and the software is implemented by VC programming. Run on the Win32 platform. In Liupanshui, a piano teacher and two students were invited to play. After comparing their different results. All the indexes of the system design have met the requirements.
【學(xué)位授予單位】：貴州大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2008
【分類號(hào)】：J624.1

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 張應(yīng)輝;;兒童鋼琴教學(xué)現(xiàn)狀思考[J];遼寧行政學(xué)院學(xué)報(bào);2006年11期

相關(guān)博士學(xué)位論文 前1條

1 劉濤;音樂情感認(rèn)知模型與交互技術(shù)研究[D];浙江大學(xué);2006年

，

本文編號(hào)：1423472