本文選題：人體動作識別 + 協(xié)同訓(xùn)練�。� 參考：《江蘇大學(xué)》2017年碩士論文

【摘要】：隨著科學(xué)技術(shù)的發(fā)展,人體動作識別逐漸成為人工智能和機(jī)器視覺領(lǐng)域一個重要的研究方向,具有廣闊的發(fā)展前景和很強(qiáng)的實(shí)用價值�？蓱�(yīng)用于日常的視頻監(jiān)控、智能醫(yī)療、運(yùn)動分析、人機(jī)智能交互等。同時,由于視頻中場景的復(fù)雜性、動作類內(nèi)變化,以及需要大量的有標(biāo)注樣本來訓(xùn)練泛化性能強(qiáng)的識別模型,這些都給人體動作識別的研究帶來了挑戰(zhàn)性。本文對人體動作識別的若干問題,特別是基于半監(jiān)督的人體動作識別進(jìn)行了較深入的研究。首先闡述了人體動作識別的選題背景與研究的目的和意義;其次概述了人體動作識別的關(guān)鍵技術(shù),如關(guān)鍵幀提取技術(shù)、特征提取技術(shù)以及人體動作識別技術(shù)等。在目前人體動作識別的理論研究基礎(chǔ)上,本文提出了基于混合式協(xié)同訓(xùn)練的人體動作識別方法和基于半監(jiān)督集成學(xué)習(xí)的人體動作識別方法,并在以上兩種算法的基礎(chǔ)上設(shè)計開發(fā)人體動作識別原型系統(tǒng),主要研究內(nèi)容如下:1)提出了基于混合式協(xié)同訓(xùn)練的人體動作識別方法。針對目前人體動作視頻中有標(biāo)記數(shù)據(jù)不足的問題,提出了一種基于混合式協(xié)同訓(xùn)練的新型人體動作識別方法。該方法利用動作識別領(lǐng)域不同類型的識別方法來構(gòu)建基分類器,并進(jìn)行迭代的相互訓(xùn)練以提高泛化性能,可以降低標(biāo)注成本并實(shí)現(xiàn)不同識別方法的優(yōu)勢互補(bǔ),進(jìn)而提高人體動作的識別精度。實(shí)驗結(jié)果表明,本文所提出的算法可以有效地識別視頻中的人體動作。2)提出了基于半監(jiān)督集成學(xué)習(xí)的人體動作識別方法。針對協(xié)同訓(xùn)練類算法隨著迭代次數(shù)的增加,基分類器的差異性會越來越小,以及迭代訓(xùn)練中產(chǎn)生的基分類器沒有被充分利用的問題。提出了基于協(xié)同訓(xùn)練和集成學(xué)習(xí)相結(jié)合的人體動作識別方法。該方法對每個基分類器設(shè)置一個集合。將基分類器迭代訓(xùn)練過程中產(chǎn)生的中間分類器加入到各自的集合中,然后利用這個集合來選擇偽標(biāo)號數(shù)據(jù)。并定義了一個基于置信度的最大證據(jù)邊緣函數(shù)來選擇偽標(biāo)號數(shù)據(jù),最終利用該算法對人體動作進(jìn)行識別。該方法能有效克服協(xié)同訓(xùn)練迭代過程中基分類器差異退化的問題,進(jìn)一步提高人體動作識別的準(zhǔn)確率。3)設(shè)計并實(shí)現(xiàn)了基于半監(jiān)督協(xié)同訓(xùn)練和集成學(xué)習(xí)的人體動作識別的原型系統(tǒng)。采用面向?qū)ο笳Z言C#和MATLAB進(jìn)行編程,通過原型系統(tǒng)的運(yùn)行測試,表明所提的方法可用于相應(yīng)的人體動作識別,并且該原型系統(tǒng)界面友好、功能齊全、可維護(hù)性好。
[Abstract]:With the development of science and technology, human motion recognition has gradually become an important research direction in the field of artificial intelligence and machine vision, which has broad development prospects and strong practical value. It can be used in daily video surveillance, intelligent medical treatment, motion analysis, human-computer intelligent interaction and so on. At the same time, due to the complexity of the scene in the video, the changes in the action class, and the need for a large number of labeled samples to train the generalized recognition model, all these bring challenges to the research of human motion recognition. In this paper, some problems of human motion recognition, especially semi-supervised human motion recognition, are studied in depth. Firstly, the background of the topic selection and the purpose and significance of the research are introduced. Secondly, the key technologies of human motion recognition, such as key frame extraction, feature extraction and human motion recognition, are summarized. Based on the theoretical research of human motion recognition, this paper proposes a human motion recognition method based on hybrid cooperative training and a semi-supervised integrated learning method. On the basis of the above two algorithms, the prototype system of human motion recognition is designed and developed. The main research contents are as follows: (1) A human motion recognition method based on hybrid cooperative training is proposed. Aiming at the shortage of tagged data in human motion video, a new human motion recognition method based on hybrid cooperative training is proposed. This method uses different recognition methods in the field of action recognition to construct the base classifier, and carries out iterative training to improve generalization performance. It can reduce the labeling cost and realize the complementary advantages of different recognition methods. And then improve the recognition accuracy of human body action. Experimental results show that the proposed algorithm can effectively recognize human motion in video. 2) A human motion recognition method based on semi-supervised integrated learning is proposed. As the number of iterations increases, the differences of base classifiers become smaller and smaller, and the basis classifiers generated in iterative training are not fully utilized. A method of human motion recognition based on cooperative training and integrated learning is proposed. The method sets a collection for each base classifier. The intermediate classifier generated in the iterative training process of the base classifier is added to the respective set, and then the pseudo-label data is selected by using this set. A maximum evidence edge function based on confidence degree is defined to select pseudo-label data. Finally, the algorithm is used to identify human actions. This method can effectively overcome the problem of the difference degradation of the base classifier in the iterative process of cooperative training. The prototype system of human motion recognition based on semi-supervised cooperative training and integrated learning is designed and implemented. Programming with object oriented languages C # and MATLAB, the test results of the prototype system show that the proposed method can be used for human body action recognition, and the prototype system has friendly interface, complete functions and good maintainability.
【學(xué)位授予單位】：江蘇大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 唐超;王文劍;李偉;李國斌;曹峰;;基于多學(xué)習(xí)器協(xié)同訓(xùn)練模型的人體行為識別方法[J];軟件學(xué)報;2015年11期

2 胡瓊;秦磊;黃慶明;;基于視覺的人體動作識別綜述[J];計算機(jī)學(xué)報;2013年12期

3 魯珂,趙繼東,葉婭蘭,曾家智;一種用于圖像檢索的新型半監(jiān)督學(xué)習(xí)算法[J];電子科技大學(xué)學(xué)報;2005年05期

4 于玲;吳鐵軍;;集成學(xué)習(xí):Boosting算法綜述[J];模式識別與人工智能;2004年01期

5 孫廣玲,唐降龍;基于分層高斯混合模型的半監(jiān)督學(xué)習(xí)算法[J];計算機(jī)研究與發(fā)展;2004年01期

6 藍(lán)金輝,馬寶華,藍(lán)天,周兆英;D-S證據(jù)理論數(shù)據(jù)融合方法在目標(biāo)識別中的應(yīng)用[J];清華大學(xué)學(xué)報(自然科學(xué)版);2001年02期

相關(guān)博士學(xué)位論文 前1條

1 任海兵;非特定人自然的人體動作識別[D];清華大學(xué);2003年

，

本文編號：1803996