Multi-objective Detection Network for Sheep Flocking Based on Lightweight Modeling
ZHANG Yujie
YANG Ruifeng
GUO Chenxia
Abstract：In order to solve the problems of multi-target detection algorithms with many parameters,complex network and large computation,which are not favorable for deployment in embedded devices,an improved lightweight multi-target detection al-gorithm YOLO-EFS is proposed.In this algorithm,EfficientFormerv2 is used as a backbone network for feature extraction as a way of reducing the number of parameters and computation in the network,and GSConv is used to enhance the nonlinear capability of the model and improve the performance of the model.GSConv to enhance the nonlinear capability of the model,and the introduction of VoV-GSCSP module to improve the model performance while streamlining the number of parameters of the model.The experi-ments use sheep image dataset and YOLOv7 as the benchmark model,the mAP,P and R values of the YOLO-EFS model network are 84.19％,86.45％and 78.16％,respectively,and the number of parameters and the number of floating-point operations have been reduced by 44％and 71％compared with that of the benchmark model,which ensures a higher detection accuracy,reduces the cost of the hardware,and is more suitable for porting to the embedded devices.
Keywords：image processingmulti-objective detectionlightweightEfficientFormerv2GSConv
Publication Date：2025-08-20
Online Publishing Date：2025-12-12(First online date of this platform, not the publication date of the document)
Pages：5( 29-33 )