CAJ | 학술논문

作为典型的模式识别任务，人脸识别有着巨大的实际应用价值与市场前景。理想环境下的人脸识别已经取得不俗成绩，然而，当所处环境变化（如姿态变换、夸张表情、阴阳脸、分辨率较低）时，识别难度增加，效果也急剧变差。与此同时，现有大多数方法无法实时（在线）完成人脸识别任务，这也限制了人脸识别技术的应用。为此，该文以深度神经网络为框架，使用大规模人脸库构造了一种新型实用的多层网络应用于大规模的人脸识别任务中并提出了一种新的搜索策略。实验结果表明，该套方法实时性好，识别率较高，是人脸识别较为理想的方案。
작위전형적모식식별임무，인검식별유착거대적실제응용개치여시장전경。이상배경하적인검식별이경취득불속성적，연이，당소처배경변화（여자태변환、과장표정、음양검、분변솔교저）시，식별난도증가，효과야급극변차。여차동시，현유대다수방법무법실시（재선）완성인검식별임무，저야한제료인검식별기술적응용。위차，해문이심도신경망락위광가，사용대규모인검고구조료일충신형실용적다층망락응용우대규모적인검식별임무중병제출료일충신적수색책략。실험결과표명，해투방법실시성호，식별솔교고，시인검식별교위이상적방안。
Being a typical pattern recognition task , face recognition has a huge market prospect and high practical value.Nowadays, it gives a promising effect when face to ideal environment .However, when face to poor environment ( e.g.pose change , exaggerated facial expressions , uneven illumination ) , the effect of most of the systems will decrease .On the other hand , most of the present methods cannot recognize human's face in real time.It also limits the application of face recognition .In this paper, we introduce a new method which using deep neural network as a framework to face recognition with a large scale dataset .Furthermore, we introduce a new image search strategy .The experimental results show that our method is an ideal one with it achieves a great real-time performance and has a high identification rate .