基于深度学习的生物资产检测模型YOLOSC
DOI:
作者:
作者单位:

北京信息科技大学智能信息处理研究所

作者简介:

通讯作者:

中图分类号:

TP391.1

基金项目:

北京市社会科学基金重点项目(21GLA007)


The Biological Asset Detection Model YOLOSC Based on Deep Learning
Author:
Affiliation:

Institute of Intelligent Information Processing, Beijing Information Science and Technology University

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    为提高生物资产监盘审计过程中盘点准确性和盘点效率,提出一种融入注意力机制和损失函数优化的生物资产检测模型YOLOSC。首先,将SENet注意力机制引入到YOLOv5s模型的主干网络中,以增强对生物资产图片中关键特征的提取能力;其次,采用CIoU作为检测框回归的损失函数,以提升训练过程中检测框的回归速度与定位精度;最后,构建了一个生物资产数据集对所提模型进行针对性训练,以提升模型检测效果。实验结果表明,相较于YOLOv5模型,YOLOSC的精确率、召回率、F1值和mAP分别提升了2.3%、2.1%、2.7%和1.6%,证明了所提出的生物资产检测模型YOLOSC的有效性。

    Abstract:

    In order to improve the inventory accuracy and inventory efficiency in the process of monitoring and auditing biological assets, a biological asset detection model YOLOSC incorporating the attention mechanism and loss function optimization is proposed. Firstly, the SENet attention mechanism is introduced into the backbone network of the YOLOv5s model in order to enhance the ability of extracting the key features in the pictures of the biological assets; secondly, the CIoU is adopted as the regression of the detection frames with the loss function to enhance the regression speed and localization accuracy of the detection frame during the training process; finally, a biological asset dataset is constructed for targeted training of the proposed model to enhance the model detection effect. The experimental results show that compared with the YOLOv5 model, the precision, recall, F1 value and mAP of YOLOSC are improved by 2.3%, 2.1%, 2.7% and 1.6%, respectively, which proves the effectiveness of the proposed biological asset detection model YOLOSC.

    参考文献
    相似文献
    引证文献
引用本文

关昆仑,朱思文,张仰森,等. 基于深度学习的生物资产检测模型YOLOSC[J]. 科学技术与工程, , ():

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-02-07
  • 最后修改日期:2024-04-02
  • 录用日期:2024-04-08
  • 在线发布日期:
  • 出版日期:
×
亟待确认版面费归属稿件,敬请作者关注