摘要
由于深度学习需要海量的数据和计算资源,对于移动设备有限的储存和计算能力来说,存在一定的技术局限性。然而利用服务端的海量数据和计算资源将模型训练完成,将训练好的模型部署到移动端,只利用移动端的计算能力是可行的。本文设计了一种基于深度学习的物体实时检测模块,并且成功将训练模型部署在移动设备的安卓系统上。该方法可以实现深度学习网络在移动设备上的快速部署。
Because deep learning requires massive amounts of data and computing resources,there are certain technical limita-tions to the limited storage and computing power of mobile devices.However,using the massive data and computing resources of the server to train the model and deploying the trained model in mobile devices,it is feasible to use only the computing power of the mobile devices.In this paper,a real-time detection module based on deep learning is designed,and the trained model is successfully deployed on android system of mobile devices.This method enables rapid deployment of deep learning networks on mobile devices.
出处
《科学技术创新》
2019年第2期76-78,共3页
Scientific and Technological Innovation
关键词
深度学习
移动设备
安卓系统
实时检测
Deep learning
Mobile device
Android system
Real-time detection
