Structural re-parameterized YOLOv5s for rapid pig detection

Mingfei Liang; Deli Zhu; Maosheng Yu

doi:10.1117/12.2673557

22 May 2023 Structural re-parameterized YOLOv5s for rapid pig detection

Mingfei Liang, Deli Zhu, Maosheng Yu

Proceedings Volume 12640, International Conference on Internet of Things and Machine Learning (IoTML 2022); 126400U (2023) https://doi.org/10.1117/12.2673557
Event: International Conference on Internet of Things and Machine Learning (IoTML 2022), 2022, Harbin, China

Abstract

To solve the problems of low detection accuracy, large amount of parameters and computation, a fast pig detection algorithm based on yolov5s is proposed in this paper. First, to enhance the learning ability of feature extraction, RepVGG Block is used to replace the ordinary convolution in the backbone network, and proposes a new feature extraction structure named R-CSPD that combined CSPDenseNet and RepVGG Block. Second, the detection speed is increased by compressing the number of channels in the neck network and using depthwise separable convolution to reduce the number of parameters and computations. Finally, C3VGG is used to optimize the C3 structure to enhance the localization ability of the target. The experimental results show that compared with the benchmark model, the calculation amount of the model in this paper is reduced by 30.7%, mAP@0.5 is increased by 1 percentage point, the mAP@0.5:0.95 is increased by 2.4 percentage points, and the detection speed is increased by 17 FPS.

Citation Download Citation

Mingfei Liang, Deli Zhu, and Maosheng Yu "Structural re-parameterized YOLOv5s for rapid pig detection", Proc. SPIE 12640, International Conference on Internet of Things and Machine Learning (IoTML 2022), 126400U (22 May 2023); https://doi.org/10.1117/12.2673557

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available