|国家科技期刊平台
首页|期刊导航|福建电脑|人脸识别在敏感人物检测中的应用算法研究

人脸识别在敏感人物检测中的应用算法研究OA

Research on the Application of Face Recognition in Sensitive Figure Detection Algorithm

中文摘要英文摘要

目前的视频发布平台上,用户上传的视频数量呈现爆发式增长.为解决视频发布平台上审核含有敏感人物的违规视频的难题,本文提出一种高效的人脸识别算法.该算法使用开源的人脸数据集WiderFace对Yolov5Face网络进行重新训练,将Yolov5Face检测到的人脸区域图像输入基于ArcFace的人脸识别模块中,进行人脸特征提取,并利用最小欧氏距离法与敏感人物人脸图像进行比对,准确找出含有敏感人物的视频.实验结果表明,该算法具有较好的准确性,可为视频发布平台提供可靠的支持.

At present,the number of videos uploaded by users on video publishing platforms is showing an explosive growth.To solve the problem of reviewing illegal videos containing sensitive individuals on video publishing platforms,this paper proposes an efficient facial recognition algorithm.This algorithm retrains the Yolov5Face network using the open-source face dataset WiderFace.The image of the face area detected by Yolov5Face is input into an ArcFace based face recognition module for facial feature extraction.The minimum Euclidean distance method is used to compare with the face image of sensitive individuals,accurately identifying videos containing sensitive individuals.The experimental results show that the algorithm has good accuracy and can provide reliable support for video publishing platforms.

陈英杰

福州大学物理与信息工程学院 福州 350000

计算机与自动化

敏感人物人脸识别YOLO目标检测ArcFace算法

Sensitive FigureFace RecognitionYOLO Object DetectionArcFace Algorithm

《福建电脑》 2024 (004)

1-4 / 4

10.16707/j.cnki.fjpc.2024.04.001

评论