Research Article
Pneumonia Image Recognition Based on Transfer Learning
@INPROCEEDINGS{10.1007/978-3-031-32443-7_8, author={Tao Zhong and HuiTing Wen and Zhonghua Cao and Xinhui Zou and Quanhua Tang and Wenle Wang}, title={Pneumonia Image Recognition Based on Transfer Learning}, proceedings={Mobile Networks and Management. 12th EAI International Conference, MONAMI 2022, Virtual Event, October 29-31, 2022, Proceedings}, proceedings_a={MONAMI}, year={2023}, month={5}, keywords={Transfer Learning pneumonia detection}, doi={10.1007/978-3-031-32443-7_8} }- Tao Zhong
HuiTing Wen
Zhonghua Cao
Xinhui Zou
Quanhua Tang
Wenle Wang
Year: 2023
Pneumonia Image Recognition Based on Transfer Learning
MONAMI
Springer
DOI: 10.1007/978-3-031-32443-7_8
Abstract
With the rapid development of artificial intelligence (AI), the anomalies detection in biomedical has became important in patients’ health monitoring. The pneumonia, including COVID-19, is a global threat. Detecting the infected patients in time is very critical to combating this epidemics. Thus, a rapid and accurate method for detecting pneumonia is urgently needed. In this paper, a deep-learning detection model, is designed to detect pneumonia efficient. Since training a neural network needs consuming a lot of time resources and computing resources, transfer learning is used for pre-training. At the same time, in order to improve the detection efficiency, we combine various deep learning models, and then perform prediction and classification. The simulation results show that comparing with the 91.5% accuracy of the traditional CNN model, the transfer learning model consisting of vgg16VGG16, vgg19VGG19, RresNnet50 and Xxecption reached 93.27%, 93.43%, 92.31% and 90.22% respectively. Most of the models are superior to the traditional models and have excellent stability with less time consuming.
