10 基于Tucker分解的音频分类方法的研究

杨立东;辛文超;胡江涛;牛大伟;张丹丹;

• 1:内蒙古科技大学

•  

摘要(Abstract)：

提出了一种基于张量分解的音频信号的分类方法.首先,通过张量建模与Tucker分解挖掘出信号潜在的结构信息,当作音频信号的有效特征信息,然后利用支持向量机作为分类器完成分类.实验结果表明:通过张量分析提取的特征具有更好的区分度,分类的正确率达到了91.3%,相较于传统方法音频分类的正确率有了明显的提高,说明张量分析方法在音频信号分类中的有效性.

关键词(KeyWords)： 音频分类;张量分析;特征提取;Tucker分解

基金项目(Foundation): 国家自然科学基金资助项目(61640012);; 内蒙古自治区自然科学基金资助项目(2017MS(LH)0602)

作者(Author): 杨立东;辛文超;胡江涛;牛大伟;张丹丹;

Email:

DOI:

参考文献(References)：

• [1] 曾毅,刘成林,谭铁牛.类脑智能研究的回顾与展望[J].计算机学报,2016(1):212.

• [2] MIYAZAK K,TODA T,HAYASHI T,et al.Environmental sound processing and its applications[J].IEEJ Transactionson Electrical and Electronic Engineering,2019,14(3).

• [3] SUN X,LIU Z,HU Y,et al.Perceptual multi-channel visual feature fusion for scene categorization[J].Information Sciences,2018,429:37.

• [4] STOWELL D.Computational bioacoustic scene analysis[M].Computational Analysis of Sound Scenes and Events.London,UK:Springer,Cham,2018:303.

• [5] AIHARA I,MIZUMOTO T,AWANO H,et al.Call alternation between specific Pairs of male frogs revealed by a sound-imaging method in their natural habitat.[J].ProcInterspeech,2016(1):2597.

• [6] JANCOVIC P,KOKUE M .Acoustic recognition of multiple bird species based on penalised maximum likelihood[J].IEEE Signal Processing Letters,2015(2):1.

• [7] 杨立东,王晶,谢湘,等.基于张量分解模型的语音信号特征提取方法[J].北京理工大学学报,2013,33(11):1171.

• [8] 顾树生,赵玲,侯军.一种基于奇异值分析的多变量系统设计方法[J].控制与决策,1993,32(11):37.

• [9] 张军委.基于音频特征融合的设备异常识别研究.[D].济南:山东大学,2019.

• [10] YANG L,WANG J,XIE X,et al.Application of tucker decomposition in speech signal feature extraction[C]// 2013 International Conference on Asian Language Processing,Urumqi,China:IEEE,2013:155.

• [11] 刘亦芃.一种关于深度学习网络结构的改进策略[D].长春:吉林大学,2019.

• [12] MAGARIňOS C,ERRO D,LOPEZ-OTERO P,et al.Language-independent acoustic cloning of HTS voices[J].Computer Speech & Language,2019,55:168.

• [13] WALDEKAR S,SAHA G.Classification of audio scenes with novel features in a fused system framework[J].Digital Signal Processing,2018,75:71.

• [14] JIANG Y,LEUNG F H F.Comparison of supervector and majority voting in acoustic scene identification[C]//2018 IEEE 23rd International Conference on Digital Signal Processing (DSP).Shanghai,China:IEEE,2018:1.

• [15] VALENTI M,TONELLI D,VESPERINI F,et al.A neural network approach for sound event detection in real life audio[C]//2017 25th European Signal Processing Conference (EUSIPCO).Kos,Greece:IEEE,2017:2754.

• [16] LECUN Y,BENGIO Y,HINTON G.Deep learning[J].Nature,2015,521(7553):436.

• [17] ANTOUN S.Biologically inspired perception for robotics in hostile environments[J].Biologically Inspired Cognitive Architectures,2016,16:1.

• [18] ZHAO M,JIAO L,MA W,et al.Classification and saliency detection by semi-supervised low-rank representation[J].Pattern Recognition,2016,51:281.

• [19] WALDEKAR S,SAHA G.Classification of audio scenes with novel features in a fused system framework[J].Digital Signal Processing,2018,75:71.