百度EasyDL人工智能技术助力高校控烟应用研究

Application Research on Assistance of Baidu EasyDL Artificial Intelligence Technology for the Tobacco Control in Colleges and Universities

文章对目前我国大学生校园吸烟较为严重的现象进行了阐述,报告了视频监控技术、红外热成像技术以及烟雾报警技术等几种常见校园控烟方法存在的弊端。针对传统控烟技术的缺陷提出了一种基于人工智能技术应用的校园监控学生吸烟解决方案,前期对吸烟者与未吸烟者大量图片的采集,训练并构建出精准度较高的监测模型以达到自动识别吸烟者的目的,通过多次验证达到了预期的控烟效果。

This paper expounds the serious phenomenon of college students'smoking on campus in China,and reports the shortcomings of several common campus smoking control methods,such as video monitoring technology,infrared thermal imaging technology and smoke alarm technology.Aiming at the defects of traditional smoking control technology,this paper proposes a solution for campus monitoring students'smoking based on the application of Artificial Intelligence technology.In the early stage,a large number of pictures of smokers and non-smokers are collected,and a monitoring model with high accuracy is trained and constructed to achieve the goal of automatic identification of smokers.The expected smoking control effect is achieved through multiple verifications.