Abstract
DNN-HMM based acoustic model for continuous pig cough sound recognition
Highlights
The global demand for meat products is expected to increase steadily[1,2,3], while pork accounts for a huge part of the meat products
Deep Belief Network (DBN) is a generative model proposed by Hinton[42], forming by Restricted Boltzmann Machines (RBM)[43]
By comparing the test results with the standard labels, we can see that the second non-pig cough sound in the recognition result is the insertion error, the third non-pig cough sound is the substitution error, and the unidentified last non-pig cough sound in the standard labels is classified as the deletion error
Summary
The global demand for meat products is expected to increase steadily[1,2,3], while pork accounts for a huge part of the meat products. Jian Zhao, et al DNN-HMM based acoustic model for continuous pig cough sound recognition. In 2012, Milone et al.[17] proposed an acoustic model based on HMM for continuous cattle ingestive sounds recognition, which divided the continuous cattle ingestive sounds into three syllables: “chews”, “bites” and “chewbites”. The HMM[21,22] is used to construct the acoustic model of continuous pig sound in farm conditions. The factors affecting pig cough and non-cough are set as hidden states of HMM, and feature vectors of continuous pig sound as observation states of HMM. DNN can be used to model feature vectors of continuous pig sound and to describe the corresponding relation between HMM observation states and hidden states. We propose a new method for continuous pig cough recognition based on DNN-HMM acoustic model[28,29,30]. The experimental results and discussion are presented in the fourth part, followed by the conclusions in the fifth part
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Agricultural and Biological Engineering
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
