基于稀疏分量分析的生猪音频欠定盲源分离研究

doi:10.11843/j.issn.0366-6964.2023.07.013

摘要/Abstract

摘要： 旨在针对生猪养殖过程中，混合生猪音频特征难以提取及识别的问题，提出一种基于稀疏化理论的欠定生猪盲源信号分离方法。本研究选取4个月、150 kg左右，健康状况良好的长白母猪，将其不同状态下的叫声按照不同系数混合得到的音频信号作为观测信号，运用短时傅里叶变换（short-time Fourier transform，STFT）对音频信号做时频域转换，通过分组筛选出信号中的单源点，使用自适应阻尼系数的AP （affinity propagation）算法结合奇异值分解，将单源点聚类以估计混合矩阵，采用优化最小l_p范数的方法完成音频信号的重构。设计2组试验，1组阐述试验的一般过程，另1组通过对比分析整个分离算法的性能，使用相似系数（similarity coefficient）、信噪比（signal to noise ratio，SNR）和均方误差（mean square error，MSE）衡量分离音频质量。结果表明：1）3个源信号与2个观测信号的欠定生猪盲源分离中，不同时长下分离出的音频信号与对应源信号的相似系数、信噪比和均方误差分别在0.67~0.92、7.9~9.7 dB和0.005~0.08之间，从波形上看，算法的分离性能与时间长短和试验次数无关，结果具有一定的稳定性。2）在源信号数与观测信号数分别为3和2、4和2、4和3、5和2、5和3、5和4时，重构信号与源信号的平均相似系数、信噪比和均方误差分别在0.785~0.957、7.468~10.347 dB和0.019~0.092之间，经过对比分析，本研究方法具有一定的可靠性。3）在源信号数一定时，观测信号数越多，测得的指标越好，分离出来的音频质量越优。综上所述，该方法能够较为有效地分离出混合猪声信号的各源信号分量，为实际环境中混合生猪音频的特征提取奠定了基础。

关键词: 生猪音频, 信号稀疏化, AP聚类, l_p范数, 盲源分离

Abstract: In the process of pig breeding, it was difficult to extract and recognize audio features of mixed pigs. A blind source signal separation method based on sparse theory was proposed for underdetermined pigs. The 4 months old and 150 kg sows in good health were selected, and the audio signals obtained by mixing their cries in different states according to different coefficients were used as observation signals. Short-time Fourier transform was used to convert the audio signals. Single source points in the signals were selected by grouping, and AP algorithm of adaptive damping coefficient combined with singular value decomposition was used. The single source points were clustered to estimate the mixing matrix, and the audio signal was reconstructed by optimizing the minimum l_p norm. Two groups of experiments were designed. One group explained the general process of the experiment, and the other group analyzed the performance of the whole separation algorithm by comparison. Similarity coefficient, signal-to-noise ratio and mean square error were used to measure the separation audio quality. The results showed that:1) In the underdetermined pig blind source separation of 3 source signals and 2 observation signals, the similarity coefficient, signal-to-noise ratio and mean square error of the separated audio signals and the corresponding source signals at different time ranged from 0.67-0.92, 7.9-9.7 dB and 0.005-0.08, respectively. From the view of waveform, the separation performance of the algorithm was independent of the length of time and the number of tests, and the results had a certain stability. 2) When the number of source signals and observed signals were 3 and 2, 4 and 2, 4 and 3, 5 and 2, 5 and 3, 5 and 4, the average similarity coefficient, signal-to-noise ratio and mean square error of reconstructed signals and source signals were 0.785-0.957, 7.468-10.347 dB and 0.019-0.092, respectively. After comparative analysis, this research method had certain reliability. 3) When the number of source signals was certain, the more the number of observed signals, the better the index measured, and the better the separated audio quality. In conclusion, this method can effectively separate each source signal component of mixed pig audio signal, which layed a foundation for feature extraction of mixed pig audio in real environment.

Key words: pig audio, signal sparsification, AP clustering, l_p norm, blind source separation

中图分类号:

TN911.7

彭硕, 陶亮, 查文文, 陈成鹏, 辜丽川, 朱诚, 焦俊. 基于稀疏分量分析的生猪音频欠定盲源分离研究[J]. 畜牧兽医学报, 2023, 54(7): 2794-2809.

PENG Shuo, TAO Liang, ZHA Wenwen, CHEN Chengpeng, GU Lichuan, ZHU Cheng, JIAO Jun. Underdetermined Blind Source Separation Algorithm for Pig Audio Signals Based on Sparse Component Analysis[J]. Acta Veterinaria et Zootechnica Sinica, 2023, 54(7): 2794-2809.

参考文献 42

[1]	冯泽猛, 张云华, 贺玉敏, 等.智能养殖:生猪行为研究及其应用[J].农业现代化研究, 2021, 42(1):1-9.FENG Z M, ZHANG Y H, HE Y M, et al.Intelligent breeding:the research and application of pig behavior[J].Research of Agricultural Modernization, 2021, 42(1):1-9.(in Chinese)
[2]	刘铃, 王丹丹, 崔凯, 等.猪繁殖与呼吸综合征抗病育种研究进展[J].畜牧兽医学报, 2023, 54(2):434-442.LIU L, WANG D D, CUI K, et al.Advances of disease-resistant breeding on porcine reproductive and respiratory syndrome[J].Acta Veterinaria et Zootechnica Sinica, 2023, 54(2):434-442.(in Chinese)
[3]	刘龙申, 舒翠霓, 李波, 等.基于EfficientDet的围产期母猪姿态识别[J].农业机械学报, 2022, 53(4):271-279.LIU L S, SHU C N, LI B, et al.Recognition of perinatal sows' posture based on EfficientDet algorithm[J].Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(4):271-279.(in Chinese)
[4]	沈明霞, 太猛, CEDRIC O, 等.基于深层卷积神经网络的初生仔猪目标实时检测方法[J].农业机械学报, 2019, 50(8):270-279.SHEN M X, TAI M, CEDRIC O, et al.Real-time detection method of newborn piglets based on deep convolution neural network[J].Transactions of the Chinese Society for Agricultural Machinery, 2019, 50(8):270-279.(in Chinese)
[5]	邱洪涛.基于深度学习与证据理论的多源生猪呼吸道疾病识别研究[D].合肥:安徽农业大学, 2019.QIU H T.Recognition of respiratory diseases in pigs with multiple sources based on deep learning and evidence theory[D].Hefei:Anhui Agricultural University, 2019.(in Chinese)
[6]	CHA C N, PARK E K, YOO C Y, et al.Changes in the appetite and behavior of cattle and pigs inoculated with the foot-and-mouth disease vaccine[J].J Prev Vet Med, 2016, 40(4):166-168.
[7]	刘振宇, 赫晓燕, 桑静, 等.基于隐马尔可夫模型的猪咳嗽声音识别的研究[C]//中国畜牧兽医学会.中国畜牧兽医学会信息技术分会第十届学术研讨会论文集.北京:中国农业大学出版社, 2015:107-112.LIU Z Y, HE X Y, SANG J, et al.Research on pig cough sound recognition based on hidden Markov model[C]//Chinese Association of Animal Science and Veterinary Medicine.Proceedings of the 10th Symposium of Information Technology Branch of Chinese Society of Animal Husbandry and Veterinary Medicine.Beijing:China Agricultural University Press, 2015:107-112.(in Chinese)
[8]	ZHAO J, LI X, LIU W H, et al.DNN-HMM based acoustic model for continuous pig cough sound recognition[J].Int J Agric Biol Eng, 2020, 13(3):186-193.
[9]	吴亚文.基于音频压缩传输与端点检测的生猪音频识别研究[D].合肥:安徽农业大学, 2021.WU Y W.Research on pig audio recognition based on audio compression transmission and endpoint detection[D].Hefei:Anhui Agricultural University, 2021.(in Chinese)
[10]	彭硕, 刘东阳, 时国龙, 等.基于深度神经网络及隐马尔科夫模型的生猪状态音频识别[J].中国农业大学学报, 2022, 27(6):172-181.PENG S, LIU D Y, SHI G L, et al.Pig state audio recognition based on deep neural network and hidden Markov model[J].Journal of China Agricultural University, 2022, 27(6):172-181.(in Chinese)
[11]	CORDEIRO A F D S, NÄÄS I D A, DA SILVA LEITÃO F, et al.Use of vocalisation to identify sex, age, and distress in pig production[J].Biosyst Eng, 2018, 173:57-63.
[12]	冷碧滨, 吉雪强, 曾颢, 等.中国大规模生猪养殖技术效率研究[J].浙江农业学报, 2018, 30(6):1082-1088.LENG B B, JI X Q, ZENG H, et al.Study on technical efficiency of large-scale pig breeding in China[J].Acta Agriculturae Zhejiangensis, 2018, 30(6):1082-1088.(in Chinese)
[13]	YANG Z Y, ZHOU G X, XIE S L, et al.Blind spectral unmixing based on sparse nonnegative matrix factorization[J].IEEE Trans Image Process, 2010, 20(4):1112-1125.
[14]	ZHOU G X, ZHAO Q B, ZHANG Y, et al.Linked component analysis from matrices to high-order tensors:applications to biomedical data[J].Proc IEEE, 2016, 104(2):310-331.
[15]	JUN H, CHEN Y, ZHANG Q H, et al.Blind source separation method for bearing vibration signals[J].IEEE Access, 2018, 6:658-664.
[16]	GEORGIEV P, THEIS F, CICHOCKI A.Sparse component analysis and blind source separation of underdetermined mixtures[J]. IEEE Trans Neural Netw, 2005, 16(4):992-996.
[17]	GUO Q, RUAN G Q, LIAO Y P.A time-frequency domain underdetermined blind source separation algorithm for MIMO radar signals[J].Symmetry, 2017, 9(7):104.
[18]	CHEN S S, DONOHO D L, SAUNDERS M A.Atomic decomposition by basis pursuit[J].SIAM Rev, 2001, 43(1):129-159.
[19]	BOFILL P, ZIBULEVSKY M.Underdetermined blind source separation using sparse representations[J].Signal Process, 2001, 81(11):2353-2362.
[20]	GEORGIEV P G, THEIS F, CICHOCKI A.Optimization algorithms for sparse representations and applications[M]//HAGER W W, HUANG S J, PARDALOS P M, et al.Multiscale Optimization Methods and Applications.New York:Springer, 2006:85-99.
[21]	ZHEN L L, PENG D Z, YI Z, et al.Underdetermined blind source separation using sparse coding[J].IEEE Trans Neural Netw Learn Syst, 2017, 28(12):3102-3108.
[22]	JOURJINE A, RICKARD S, YILMAZ O.Blind separation of disjoint orthogonal signals:demixing N sources from 2 mixtures[C]//2000 IEEE International Conference on Acoustics, Speech, and Signal Processing.Istanbul:IEEE, 2000:2985-2988.
[23]	XIE Y, XIE K, XIE S L.Underdetermined blind separation of source using lp-norm diversity measures[J]. Neurocomputing, 2020, 411:259-267.
[24]	ARBERET S, GRIBONVAL R, BIMBOT F.A robust method to count and locate audio sources in a multichannel underdetermined mixture[J].IEEE Trans Signal Process, 2010, 58(1):121-133.
[25]	HASSAN N, RAMLI D A.Sparse component analysis (SCA) based on adaptive time-frequency thresholding for underdetermined blind source separation (UBSS)[J].Sensors, 2023, 23(4):2060.
[26]	于和新.欠定盲源分离的预估分离两步法[D].大连:大连理工大学, 2021.YU H X.Two-stages predictive and separation method for underdetermined blind source separation[D].Dalian:Dalian University of Technology, 2021.(in Chinese)
[27]	LI Y Q, CICHOCKI A, AMARI S I.Analysis of sparse representation and blind source separation[J].Neural Comput, 2004, 16(6):1193-1234.
[28]	高鹰, 谢胜利, 许若宁, 等.基于粒子群优化算法的稀疏信号盲分离[J].系统仿真学报, 2006, 18(8):2264-2266.GAO Y, XIE S L, XU R N, et al.Blind sparse source separation based on particle swarm optimization[J].Journal of System Simulation, 2006, 18(8):2264-2266.(in Chinese)
[29]	田旭, 马晓川, 封超, 等.基于稀疏分量分析的振动信号源识别方法[J].信号处理, 2021, 37(6):1034-1045.TIAN X, MA X C, FENG C, et al.Vibration signal sources identification method based on sparse component analysis[J].Journal of Signal Processing, 2021, 37(6):1034-1045.(in Chinese)
[30]	FREY B J, DUECK D.Clustering by passing messages between data points[J].Science, 2007, 315(5814):972-976.
[31]	DUAN Y X, LIU C Y, LI S, et al.An automatic affinity propagation clustering based on improved equilibrium optimizer and t-SNE for high-dimensional data[J].Inf Sci, 2023, 623:434-454.
[32]	王开军, 张军英, 李丹, 等.自适应仿射传播聚类[J].自动化学报, 2007, 33(12):1242-1246.WANG K J, ZHANG J Y, LI D, et al.Adaptive affinity propagation clustering[J].Acta Automatica Sinica, 2007, 33(12):1242-1246.(in Chinese)
[33]	CHARTRAND R, YIN W T.Iteratively reweighted algorithms for compressive sensing[C]//2008 IEEE International Conference on Acoustics, Speech and Signal Processing.Las Vegas:IEEE, 2008:3869-3872.
[34]	SAAB R, YILMAZ O, MCKEOWN M J, et al.Underdetermined anechoic blind source separation via lq-basis-pursuit with q
[35]	YANG J J, GUO Y, YANG Z Y, et al.Under-determined convolutive blind source separation combining density-based clustering and sparse reconstruction in time-frequency domain[J].IEEE Trans Circuits Syst I Regul Pap, 2019, 66(8):3015-3027.
[36]	SAWADA H, ARAKI S, MAKINO S.Underdetermined convolutive blind source separation via frequency bin-wise clustering and permutation alignment[J].IEEE Trans Audio Speech Lang Process, 2011, 19(3):516-527.
[37]	HE X S, HE F, CAI W H.Underdetermined BSS based on K-means and AP clustering[J].Circuits Syst Signal Process, 2016, 35(8):2881-2913.
[38]	季策, 姜雨田.基于方向幅值比的欠定盲源分离算法[J].东北大学学报:自然科学版, 2019, 40(7):920-924.JI C, JIANG Y T.Underdetermined blind source separation algorithm based on directional amplitude ratio[J].Journal of Northeastern University:Natural Science, 2019, 40(7):920-924.(in Chinese)
[39]	马宝泽, 张天骐, 安泽亮, 等.一种多维信源衰减延时混合的欠定盲源分离方法[J].电子与信息学报, 2021, 43(8):2258-2266.MA B Z, ZHANG T Q, AN Z L, et al.An underdetermined blind source separation approach for attenuated and time-delayed mixtures of multiple sources[J].Journal of Electronics & Information Technology, 2021, 43(8):2258-2266.(in Chinese)
[40]	马宝泽.基于瞬时和卷积混合模型的盲信号分离方法研究[D].重庆:重庆邮电大学, 2021.MA B Z.A dissertation submitted to chongqing university of posts and telecommunications in partial fulfillment of the requirement for the degree of doctor of engineering[D].Chongqing:Chongqing University of Posts and Telecommunications, 2021.(in Chinese)
[41]	解元, 邹涛, 孙为军, 等.面向高混响环境的欠定卷积盲源分离算法[J].通信学报, 2023, 44(2):82-93.XIE Y, ZOU T, SUN W J, et al.Algorithm of underdetermined convolutive blind source separation for high reverberation environment[J]. Journal on Communications, 2023, 44(2):82-93.(in Chinese)
[42]	解元, 邹涛, 孙为军, 等.面向卷积混叠环境下的盲源分离新方法[J/OL].自动化学报, 2023, doi:10.16383/j.aas.c211207.XIE Y, ZOU T, SUN W J, et al.Novel blind source separation method for convolutive mixed environment[J/OL].Acta Automatica Sinica, 2023, doi:10.16383/j.aas.c211207.(in Chinese)