首页|基于CNN的人脸年龄与性别检测系统设计

基于CNN的人脸年龄与性别检测系统设计

扫码查看
基于CNN的人脸年龄与性别检测系统以PyTorch中ResNet-50模型为主干网络,利用Wiki平台进行人脸原始图像采集预处理,构建年龄预测与性别判断分类器作为模型网络输出层,使用均方误差与交叉熵损失函数对年龄与性别结果进行预测输出.系统前端使用PyQt5构建图像识别、视频识别和实时摄像识别的交互界面,采用OpenCV对图像、视频或摄像头读取的每一帧图像进行人脸标注,将系统模型识别的人脸性别与年龄预测检测结果绘制成图像并展示.
Design of facial age and gender detection system based on CNN
The ResNet-50 model in Python is used as the backbone network for the facial age and gender detection system based on CNN,the original facial image collection is preprocessed on Wiki platform.An age prediction and gender judgment classi-fier is constructed as the output layer of the model network,and the mean square error and cross entropy loss function are used to predict and output the results of age and gender.An interactive interface of image,video and real-time camera recognition is built on front-end system used PyQt5,and OpenCV is used to annotate the face of each frame of images,videos,or camera readings.The predicted detection results of facial gender and age recognized by the model system are plotted and displayed as the image.

CNNResNet-50 modelOpenCVface recognition

孙小广、万若楠、余光正

展开 >

广州城市理工学院电子信息工程学院,广州 510800

华南理工大学物理与光电学院声学研究所,广州 510640

CNN ResNet-50模型 OpenCV 人脸识别

国家自然科学基金项目

12074129

2024

现代计算机
中大控股

现代计算机

影响因子:0.292
ISSN:1007-1423
年,卷(期):2024.30(4)
  • 9