计算机应用 ›› 2019, Vol. 39 ›› Issue (8): 2402-2407.DOI: 10.11772/j.issn.1001-9081.2019010133

• 虚拟现实与多媒体计算 • 上一篇    下一篇

基于多尺度双线性卷积神经网络的多角度下车型精细识别

刘虎, 周野, 袁家斌   

  1. 南京航空航天大学 计算机科学与技术学院, 南京 210000
  • 收稿日期:2019-02-13 修回日期:2019-04-03 发布日期:2019-04-17 出版日期:2019-08-10
  • 通讯作者: 周野
  • 作者简介:刘虎(1974-),男,陕西兴平人,高级工程师,博士,主要研究方向:物联网、云计算、智能视频处理、大数据分析;周野(1993-),男,江苏东海人,硕士研究生,主要研究方向:大数据、数据挖掘、计算机视觉;袁家斌(1968-),男,江苏泰兴人,教授,博士,主要研究方向:高性能计算。
  • 基金资助:
    江苏省产学研前瞻性联合研究项目(BY2016003-11)。

Fine-grained vehicle recognition under multiple angles based on multi-scale bilinear convolutional neural network

LIU Hu, ZHOU Ye, YUAN Jiabin   

  1. College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing Jiangsu 210000, China
  • Received:2019-02-13 Revised:2019-04-03 Online:2019-04-17 Published:2019-08-10
  • Supported by:
    This work is partially supported by the Prospective Joint Research Project of Industry, University and Research in Jiangsu Province (BY2016003-11).

摘要: 针对多角度下车辆出现一定的尺度变化和形变导致很难被准确识别的问题,提出基于多尺度双线性卷积神经网络(MS-B-CNN)的车型精细识别模型。首先,对双线性卷积神经网络(B-CNN)算法进行改进,提出MS-B-CNN算法对不同卷积层的特征进行了多尺度融合,以提高特征表达能力;此外,还采用基于中心损失函数与Softmax损失函数联合学习的策略,在Softmax损失函数基础上分别对训练集每个类别在特征空间维护一个类中心,在训练过程中新增加样本时,网络会约束样本的分类中心距离,以提高多角度情况下的车型识别的能力。实验结果显示,该车型识别模型在CompCars数据集上的正确率达到了93.63%,验证了模型在多角度情况下的准确性和鲁棒性。

关键词: 车型精细识别, 卷积神经网络, 双线性卷积神经网络, 中心损失, 多尺度

Abstract: In view of the problem that it is difficult to accurately recognize the type of vehicle due to scale change and deformation under multiple angles, a fine-grained vehicle recognition model based on Multi-Scale Bilinear Convolutional Neural Network (MS-B-CNN) was proposed. Firstly, B-CNN was improved and then MS-B-CNN was proposed to realize the multi-scale fusion of the features of different convolutional layers to improve feature expression ability. In addition, a joint learning strategy was adopted based on center loss and Softmax loss. On the basis of Softmax loss, a category center was maintained for each category of the training set in the feature space. When new samples were added in the training process, the classification center distances of samples were constrained to improve the ability of vehicle recognition in multi-angle situations. Experimental results show that the proposed vehicle recognition model achieved 93.63% accuracy on CompCars dataset, verifying the accuracy and robustness of the model under multiple angles.

Key words: fine-grained vehicle recognition, Convolutional Neural Network (CNN), Bilinear Convolutional Neural Network (B-CNN), center loss, multi-scale

中图分类号: