Details
Presenter(s)
![Bingqiang Liu Headshot](https://confcats-catavault.s3.amazonaws.com/CATAVault/ieeecass/master/files/styles/cc_user_photo/s3/user-pictures/21392_0.jpg?h=f055a199&itok=coOYxo55)
Display Name
Bingqiang Liu
- Affiliation
-
AffiliationHuazhong University of Science and Technology
- Country
Abstract
We propose an energy-efficient intelligent pulmonary auscultation system for post COVID-19 era wearable monitoring, which can be deployed on wearable devices for pre-diagnosis. Specifically, a tightly coupled two-stage hybrid neural network model with a multi-task training method is proposed to perform two-category coarse classification of normal and abnormal lung sound at the first stage, and then perform four-category classification at the second stage only if the lung sound is abnormal. Advanced lightweight CNN (convolutional neural network) structures are used to improve the performance of the model and reduce the model’s computation as well as the required power consumption.