许艇,
冯定忠,
蒋美仙,,
吴光华
浙江工业大学机械工程学院? ?杭州? ?310023
基金项目:国家自然科学基金(51605442)，浙江省科技厅公益项目(LGN18G010002)
详细信息
作者简介:张烨：男，1973年生，副教授，硕士生导师，研究方向为物联网、深度学习、无线传感器网络的设计与仿真等
许艇：男，1993年生，硕士生，研究方向为计算机视觉、深度学习、物联网技术等
冯定忠：男，1963年生，教授，博士生导师，研究方向为企业智能物流、工业工程技术及应用等
蒋美仙：女，1973年生，副教授，硕士生导师，研究方向为企业物流、系统工程等
吴光华：男，1983年生，讲师，博士，研究方向为智能物流、物联网技术等
通讯作者:蒋美仙 1056294025@qq.com
中图分类号:TP391.41计量
文章访问数:1906
HTML全文浏览量:1365
PDF下载量:65
被引次数:0
出版历程
收稿日期:2018-07-13
修回日期:2019-01-28
网络出版日期:2019-02-18
刊出日期:2019-06-01
Research on Faster RCNN Object Detection Based on Hard Example Mining
Ye ZHANG,Ting XU,
Dingzhong FENG,
Meixian JIANG,,
Guanghua WU
College of Mechanical Engineering, Zhejiang University of Technology, Hangzhou 310023, China
Funds:The National Natrual Science Foundation of China (51605442), Science Technology Department of Zhejiang Province (LGN18G010002)
摘要
摘要:针对经典的快速区域卷积神经网络(Faster RCNN)训练过程存在太多难训练样本、召回率低等问题，该文采用一种基于在线难分样本挖掘技术(OHEM)与负难分样本挖掘(HNEM)技术相结合的方法，通过训练中实时筛选的最大损失值难分样本进行误差传递，解决了模型对难分样本检测率低问题，提高模型训练效率；为更好地提高模型的召回率和模型的泛化性，该文改进了非极大值抑制(NMS)算法，设置了置信度阈值罚函数，又引入多尺度、数据增强等训练方法。最后通过比较改进前后的结果，经敏感性实验分析表明，该算法在VOC2007数据集上取得了较好效果，平均精度均值从69.9%提升到了74.40%，在VOC2012上从70.4%提升到79.3%，验证了该算法的优越性。
关键词:多目标检测/
在线样本挖掘/
负难分样本挖掘/
深度学习/
非极大值抑制
Abstract:Because of the classic Faster RCNN training proccess with too many difficult training samples and low recall rate problem, a method which combines the techniques of Online Hard Example Mining (OHEM) and Hard Negative Example Mining (HNEM) is adopted, which carries out the error transfer for the difficult samples using its corresponding maximum loss value from real-time filtering. It solves the problem of low detection of hard example and improves the efficiency of the model training. To improve the recall rate and generalization of the model, an improved Non-Maximum Suppression (NMS) algorithm is proposed by setting confidence thresholds penalty function; In addition, multi-scale training and data augmentation are also introduced. Finally, the results before and after improvement are compared: Sensibility experiments show that the algorithm achieves good results in VOC2007 data set and VOC2012 data set, with the mean Average Percision (mAP) increasing from 69.9% to 74.40%, and 70.4% to 79.3% respectively, which demonstrates strongly the superiority of the algorithm.
Key words:Multiple object detection/
Online Hard Example Mining (OHEM)/
Hard Negative Example Mining (HNEM)/
Deep learning/
Non-Maximum Suppression (NMS)
PDF全文下载地址:
https://jeit.ac.cn/article/exportPdf?id=acd16269-7704-4c66-90e1-08c7d6b69d7d
