摘要
虽然目前用于人脸特征识别的方法能够较好地捕捉、分析和确定人脸的特征,呈现相对清晰的人脸特征,但是大多数方法存在人脸图像捕捉角度方面的要求,对侧面等非正面化图像的分析和特征提取不够准确和完整。基于复杂场景下人脸特征识别的需要,有必要探索多角度、综合性的人脸特征识别方法,以更好地呈现人的脸部特征。笔者在介绍人脸特征识别的基础上,着重对基于深度学习的人脸特征识别方法的应用情况以及应用中需要注意的问题进行研究,旨在为提高人脸特征识别技术水平和效果提供相应的参考。
Although the current methods for face feature recognition can better capture, analyze and determine the features of the face, and present relatively clear face features, however, most of the methods have requirements for the angle of face image capture, and are not suitable for non-face images such as sides. The analysis and feature extraction of the frontalized images is not accurate and complete.Based on the needs of facial feature recognition in complex scenes, it is necessary to explore multi-angle and comprehensive facial feature recognition methods to better present human facial features. Based on the general introduction of face feature recognition, the article focuses on the application of deep learning-based face feature recognition methods and the problems that need attention in the application, aiming to improve the level and effect of face feature recognition technology. Provide appropriate references.
作者
王一帆
陈骁
陈鑫
梁旭
WANG Yifan;CHEN Xiao;CHEN Xin;LIANG Xu(China Shipbuilding Industry System Engineering Research Institute,Beijing 100094,China;China Ship Building IT Co.,Ltd.,Beijing 100861,China)
出处
《信息与电脑》
2022年第2期78-80,共3页
Information & Computer
关键词
深度学习
人脸特征
识别方法
deep learning
facial features
recognition method
