{"title":"Handwritten Digital Detection Based on Tensorflow Building SSD Model","authors":"Sen Feng","doi":"10.1109/ICIASE45644.2019.9074012","DOIUrl":null,"url":null,"abstract":"A computer vision recognition model based on real-time detection is built, and relevant tests are made by using the theory of deep learning. The SSD convolution neural network model is built on the Tensorflow platform, and it is used for the recognition and detection of handwritten numerals. The method of making the data set which can be used to convert MNIST into SSD is given, and the training flow is given, and the experimental results are analyzed. After 50000 training, the recognition accuracy reaches 99.19%, and the location accuracy reaches 99.99%, and the recognition effect is good.","PeriodicalId":206741,"journal":{"name":"2019 IEEE International Conference of Intelligent Applied Systems on Engineering (ICIASE)","volume":"159 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE International Conference of Intelligent Applied Systems on Engineering (ICIASE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICIASE45644.2019.9074012","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
A computer vision recognition model based on real-time detection is built, and relevant tests are made by using the theory of deep learning. The SSD convolution neural network model is built on the Tensorflow platform, and it is used for the recognition and detection of handwritten numerals. The method of making the data set which can be used to convert MNIST into SSD is given, and the training flow is given, and the experimental results are analyzed. After 50000 training, the recognition accuracy reaches 99.19%, and the location accuracy reaches 99.99%, and the recognition effect is good.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
