In order to meet the needs of real-time monitoring on mobile devices, the message have observed the amazing achievements of VGG16 in the 2014 Imagenet image classification competition and the application advantages of convolutional neural network in various fields such as image classification, image segmentation and target detection. At the same time, the demand for neural network computing speed and robustness is greatly improved. However, most neural networks have a large number of parameters. In this project, the NanoDet model is trained in the Vehicles-Openimages data set, and finally deployed to the mobile terminal to complete the identification of vehicle target detection. The significance of this project is to deploy the small model with high precision and speed to the mobile terminal equipment, so that the smart phone can carry out real-time vehicle target detection. Experiments have proved that the operation speed and accuracy of the lightweight model in the mobile terminal are relatively ideal. Under the premise of high accuracy, the network size is compressed, and the test is carried out in the equipment with poor performance such as the mobile terminal. The faster calculation speed also completes the goal of real-time detection.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    A Lightweight Neural Network for Vehicle Object Detection


    Contributors:
    Sun, Xuechen (author) / Xia, Tianrui (author) / Wang, Tianyu (author) / Xu, Chunyi (author)


    Publication date :

    2022-10-12


    Size :

    1243136 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Application of Lightweight Object Detection Network in Cucumber Leaf Detection

    Li, Xia / Yue, Zhenchao / Su, Junhao et al. | British Library Conference Proceedings | 2022




    Object Detection Method of Autonomous Vehicle Based on Lightweight Deep Learning

    Guo, Rong / Xie, Xin | British Library Conference Proceedings | 2021