基于卷积神经网络和关键词的目标检测系统

An Object Detection System based on Convolutional Neural Network and Keywords

该文以互联网上的图片数据为训练数据,结合使用深度学习技术,实现了一种基于卷积神经网络和检测关键词的目标检测系统。系统根据用户输入的检测关键词,利用网络爬虫技术搜集图像用于训练。对每张图像,使用预训练的神经网络模型提取图像特征,并采用金字塔池化技术得到图像表达向量。系统基于图像表达向量学习分类器,并利用分类器对用户上传的图片进行目标检测。系统的实现具有实际意义,理想情况下可对任意具体目标实现检测(如行人、车辆、动物等)。

This paper proposes a novel object detection system,which utilizes the images stored on the Internet as training images,and adopts the deep learning technology to achieve the object detection tasks.According the keywords typed by users,this system runs the web crawlers to gather the images related to the keywords.For each image,a pre-trained convolutional neural network model is used to extract the features from this image,and then the pyramid pooling technology is applied to generate the image rep⁃resentation vector of this image.This system learns a binary classifier on the image representation vectors,and uses it to detect the objects in the images uploaded by users.Ideally,this system is capable of detecting any object such as pedestrian,car,animal and so on.