期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：4
页码：7895
DOI：10.15680/IJIRCCE.2017.05040246
出版社：S&S Publications
摘要：Infant crying can be considered a biological alarm system, and it is the first means of communication fornewborns. Infant crying signals distress or needs, calls for the attention of parents or caregivers and motivates them toalleviate the distress. Here, we mainly focused on automation of Infant’s Cry. For this implementation we use LinearFrequency Cestrum Coefficient Cepstrum (LFCC) for feature extraction and VQ codebook for matching samples usingLBG algorithm. The baby crying samples collected from various crying baby having 0-6months age. There are 150babies’ sound as training data, each of which represents the 30 hungry infant cries, 30 sleepy infant cries, 30 wanted toburp infant cries, 30 inpain infant cries, and 30 uncomfortable infant cries (could be because his/her diaper iswet/toohot/cold air or anything else).The testing data is 40, respectively 8 infant cries for each type of infant cry. Theidentification of infant cries based the minimum distance of Euclidean distance. The, classification of the cry in fiveclasses neh – hunger own – sleepy, heh – discomfort, eair –lower gas, eh – burp.Here for classification of the cry oursystem is divided into two phases. First, in training phase, in which LFCC is applied for feature extraction, and thenVQ codebooks are generated to compress the feature vectors. Second, is the testing phase in which features extractionand codebookgeneration of samples are repeated. Here, comparison of the codebook template of samples to the all theavailable templates in the database are carried based on Euclidian distance between them.LFCC effectively capture thelower as well as higher frequency characteristics than MFCC,hence we will get good results over MFCC.
