计算机工程与应用 ›› 2024, Vol. 60 ›› Issue (8): 267-273.DOI: 10.3778/j.issn.1002-8331.2302-0056

• 图形图像处理 • 上一篇    下一篇

面向目标检测的数据集标签遗漏的协同修正技术

周定威,扈静,张良锐,段飞亚   

  1. 合肥工业大学 机械工程学院,合肥 230009
  • 出版日期:2024-04-15 发布日期:2024-04-15

Collaborative Correction Technology of Label Omission in Dataset for Object Detection

ZHOU Dingwei, HU Jing, ZHANG Liangrui, DUAN Feiya   

  1. School of Mechanical Engineering, Hefei University of Technology, Hefei 230009, China
  • Online:2024-04-15 Published:2024-04-15

摘要: 针对图像标注中疲劳、粗心等因素引起的标签遗漏现象,使得模型训练时难以正确区分正负样本,进而影响模型性能。设计了一种协同修正技术,通过多次迭代更新训练集,将潜在无标签的目标进行对象擦除,降低训练集的错误监督信息,避免人工的重复检查和重复标注。该方法无需进行算法参数调整、不依赖具体网络结构,低成本地减少数据集错误从而提升模型训练精度。在YOLOv5算法的实验基础上表明协同修正操作仅迭代1次即有明显的改善效果,并在多个公共数据集上能够提升0.4%~1.4%的检测精度,当数据集中的标签遗漏率达到40%时依然能够生效。该方法对数据集中样本的数据量和类别数没有限制,可应用于电商、遥感、通用等多种目标检测场景,保持着较好的鲁棒性和泛化性。

关键词: 协同修正, 标签遗漏, 数据集优化, 目标检测, 深度学习

Abstract: For the label omission caused by fatigue, carelessness and other factors in image labeling, it is difficult to correctly distinguish positive and negative samples during model training, thus affecting the performance of the model. A collaborative correction technology is designed to update the training set through multiple rounds of iteration, erase the potential unlabeled object, reduce the error monitoring information of the training set, and avoid manual repeated inspection and labeling. This method does not need to adjust the algorithm parameters, does not depend on the specific network structure, and reduces the dataset errors at low cost to improve the model training accuracy. Based on the experiment of YOLOv5 algorithm, it is shown that the cooperative correction operation can improve the detection accuracy by 0.4%~1.4% on multiple common datasets after only one iteration, and it still takes effect when the label omission rate in the dataset reaches 40%. This method has no limit on the amount of data and the number of categories of samples in the dataset, and can be applied to multiple target detection scenarios such as e-commerce, remote sensing, general purpose, etc., maintaining good robustness and generalization.

Key words: collaborative correction, label omission, dataset optimization, object detection, deep learning