西安电子科技大学学报 ›› 2021, Vol. 48 ›› Issue (3): 85-90.doi: 10.19665/j.issn1001-2400.2021.03.011

• 计算机科学与技术&人工智能 • 上一篇    下一篇

目标检测的多尺度定位提升算法

王平1(),江雨泽2(),赵光辉2()   

  1. 1.天津航天中为数据系统科技有限公司,天津 300301
    2.西安电子科技大学 人工智能学院,陕西 西安 710071
  • 收稿日期:2019-12-02 出版日期:2021-06-20 发布日期:2021-07-05
  • 通讯作者: 江雨泽
  • 作者简介:王 平(1981—),男,高级工程师,E-mail:wangping@spacezw.com|赵光辉(1980—),男,教授,博士,E-mail:ghzhao@xidian.edu.cn
  • 基金资助:
    科技部国家重点研发计划(2017YFC1404900)

Object detection based on the multiscale location Enhancement network

WANG Ping1(),JIANG Yuze2(),ZHAO Guanghui2()   

  1. 1. Tianjin Zhongwei Aerospace Data System Technology Co.,Ltd,Tianjing 300301,China
    2. School of Artificial Intelligence,Xidian University,Xi’an 710071,China
  • Received:2019-12-02 Online:2021-06-20 Published:2021-07-05
  • Contact: Yuze JIANG

摘要:

对于目标检测任务,深度神经网络模型中的一阶段网络结构存在两个问题。首先,网络结构中的锚框超参数设计的合适与否将影响整个网络的训练结果;其次,较大的降采样因子会影响目标的定位能力。针对这两个问题,提出了多尺度定位提升网络模型。重新设计了一阶段网络模型结构,并且提出了更好的锚框超参数选择方案,它在保证一阶段网络效率的同时,定位精度比之前的一阶段网络模型更好。大量实验表明,多尺度定位提升算法在保证实时性的同时实现了更高的定位精度,在公开数据集(Pascal VOC 2007)上实现了82.5%的平均准确率。

关键词: 神经网络, 多尺度定位, 目标检测, 卷积神经网络

Abstract:

For the target detection task,there are two problems in the one-stage network structure of the deep neural network model.First,whether the design of the anchor box hyperparameter is suitable or not will affect the training results of the whole network;second,a large down sampling factor will affect the positioning ability of the target.To solve these problems,this paper proposes a multi-location enhancement network.The structure of the one-stage network model is redesigned,and a better scheme for selecting the super parameters of the anchor frame is proposed.So the efficiency of the first stage network is ensured and the positioning accuracy is better than the previous one.A large number of experiments show that the multi-location enhancement network can achieve a higher positioning accuracy while ensuring real-time performance.The average accuracy of 82.5 is achieved on the public dataset (Pascal VOC 2007).

Key words: neural networks, multiscalepositioning, objectdetection, convolutional neural network

中图分类号: 

  • TN957.52