基于BiLSTM的生猪音频识别Pig Audio Recognition Based on BiLSTM
邵睿;彭硕;查文文;陈成鹏;辜丽川;焦俊;
摘要(Abstract)：
以5种猪声为研究对象，首先，用维纳滤波和端点检测对猪声进行预处理，获得有效语料；然后，提取梅尔倒谱系数(Mel Frequency Cepstral Coefficient, MFCC)制作样本集，再构建基于BiLSTM的声学模型学习样本集；最后用训练好的模型对猪声MFCC序列进行分类，实现生猪音频识别。结果表明：(1)通过5折交叉试验验证，5组模型总体识别率均达到90%,最高组为92.52%;(2)用样本集外语料对最优组模型进行算法应用测试，模型对进食、咳嗽、发情、嚎叫和哼叫的样本识别率分别为88.35%、93.65%、90.38%、88.46%、92.63%,总体识别率为90.70%。
关键词(KeyWords)： 维纳滤波;梅尔倒谱系数;双向长短时记忆网络;声学模型;加权交叉熵函数
基金项目(Foundation): 安徽省科技重大专项“大数据环境下的生猪健康养殖与疫病防控预警关键技术研究”(201903a06020009);; 安徽农业大学研究生创新基金项目“基于改进Bi-LSTM的生猪音频分类模型”(2021yjs-52)资助
作者(Authors): 邵睿;彭硕;查文文;陈成鹏;辜丽川;焦俊;
参考文献(References)：
- [1] 吴亚文，赵家龙，侯金波，等.基于改进压缩感知的生猪音频信号处理[J].江苏农业科学，2020,48(10):250-255.
- [2] Ferrari S,Silva M,Guarino M,et al.Cough Sound Analysis to Identify Respiratory Infection in Pigs[J].Computers and Electronics in Agriculture,2008,64:318-325.
- [3] Exadaktylos V,Silva M,AertsJ-M,et al.Real-time Recognition of Sick Pig Cough Sounds[J] .Computers and Electronics in Agriculture,2008 (2),63:207-214.
- [4] 徐亚妮，沈明霞，闫丽，等.待产梅山母猪咳嗽声识别算法的研究[J].南京农业大学学报，2016,39(4):681-687.
- [5] Chung Y,Oh S,Lee J,et al.Automatic Detection and Recognition of Pig Wasting Diseases Using Sound Data in Audio Surveilance Systems[J].Sensors,2013,13(10):12929-12942.
- [6] 韩磊磊.基于支持向量机的生猪异常声音识别与监测系统[D].太原：太原理工大学电气与动力工程学院，2019.
- [7] 黎煊，赵建，高云，等.基于深度信念网络的猪咳嗽声识别[J].农业机械学报，2018,49(3):179-186.
- [8] Zhao Jian,Li Xuan,Liu Wanghong,et al.DNN-HMM Based Acoustic Model for Continuous Pig Cough Sound Recognition[J].IJABE,2020,13(3):8-15.
- [9] 苍岩，罗顺元，乔玉龙.基于深层神经网络的猪声音分类[J].农业工程学报，2020,36(9):195-204.
- [10] 黎煊，赵建，高云，等.基于连续语音识别技术的猪连续咳嗽声识别[J].农业工程学报，2019,35(6):174-180.
- [11] Van HirtumA,BerckmansD.Considering the Influence of Artificial Environmental Noise to Study Cough Time-frequency Features[J].Journal of Sound and Vibration ,2003,266(3),667-675.
- [12] 张亮，龚卫国.一种改进的维纳滤波语音增强算法[J].计算机工程与应用，2010,46(26):129-131.
- [13] 吴亚文，邵睿，李淼，等.基于改进EMD-TEO倒谱距离的生猪音频信号端点检测[J].中国农业大学学报，2021,26(4):104-116.
- [14] Williams R,Zipser D.A.Learning Algorithm for Continually Running Fully Recurrent Neural Networks[J].Neural Computation,1989,1 (2),270–280.
- [15] Hochreiter S,Schmidhuber J.Long Short-term Memory[J].Neural Computation,1997,9 (8):1735-1780.
- [16] Graves A.Supervised Sequence Labelling with Recurrent Neural Networks[M].Berlin:Springer-Verlag,2012.
- [17] Bengi Y,Simard P,Frasconi P.Learning Long-term Dependencies with Gradient Descent is Different is Difficult [C]// IEEE Transactions on Neural Networks,IEEE,1994,5(2):157-166.
- [18] Hochreiter S.The Vanishing Gradient Problem During Learning Recurrent Neural Nets and Problem Solution[J].International Journal of Uncertainty Fuzziness and Knowledge-Based Systems,1998,6(2):107-116.
- [19] Schuster M,Paliwai K K.Bidirectional Recurrent Neural Networks[J].IEEETranscations on Signal Processing,2002,45(11):2673-2681.
- [20] Chen K,Huo Q.Training Deep Bidirectional LSTM Acoustic Model for LVCSR by a Context-Sensitive-Chunk BPTT Approach[J].IEEE/ACM Transactions on Audio,Speech,and Language Processing,2016,24(7):1185-1193.
- [21] Srivastava N,Hinton G,Krizhevsky A,et al.Dropout:A Simpleway to Prevent Neural Networks from Overfitting[J].Journal of Machine Learning Research,2014,15(1):1929-1958.