Education, Science, Technology, Innovation and Life
Open Access
Sign In

A Medical Waste Classification Method Based on Yolov4

Download as PDF

DOI: 10.23977/jaip.2022.050206 | Downloads: 3 | Views: 601

Author(s)

Yupeng Mou 1, Liyan Zhang 1

Affiliation(s)

1 School of Computer & Telecommunication Engineering, Dalian Jiaotong Unniversity, Huanghe Road, Dalian, China

Corresponding Author

Liyan Zhang

ABSTRACT

In recent years, with the outbreak of corona virus, medical waste will also become a way of virus transmission. Therefore, this paper presents a medical waste classification method based on improved YoloV4. This method uses hard mish activation function to replace mish activation function on the basis of YoloV4 algorithm, which reduces the amount of calculation and improves the classification speed. Two CBAM Block attention mechanism module are added, and the experimental results verify the effectiveness of this algorithm.

KEYWORDS

YoloV4, CBAM, Classification

CITE THIS PAPER

Yupeng Mou, Liyan Zhang, A Medical Waste Classification Method Based on Yolov4. Journal of Artificial Intelligence Practice (2022) Vol. 5: 42-46. DOI: http://dx.doi.org/10.23977/jaip.2022.050206.

REFERENCES

[1] L.Y. Zhang, Y.X. Zhao, Y.P. Mou, et al. Automatic classification algorithm of train common garbage based on resnet50, Journal of Dalian Jiaotong University, 2021, 42 (04): 101-105
[2] J Redmon, S Divvala, R Girshick, et al. You only look once: Unified, real-time object detection, IEEE conference on computer vision and pattern recognition. 2016: 779-788.
[3] X.T. Du, L.Y. Zhang, Y.X. Zha, et al. Vehicle distance measurement algorithm based on Yolov3, Computer programming skills and maintenance, 2020, (04): 15-16.
[4] P.Ramachandran, B. Zoph, Q.V.Le. Searching for activation functions. arXiv preprint arXiv:1710.05941, 2017.
[5] A.Howard, M.Sandler, G. Chu, et al. Searching for mobilenetv3. the IEEE/CVF international conference on computer vision. 2019: 1314-1324.

Downloads: 5240
Visits: 171846

Sponsors, Associates, and Links


All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.