Brightness and Contrast Adaptive Face Recognition System

2023 International Technical Conference on Circuits/Systems, Computers, and Communications (ITC-CSCC)(2023)

引用 0|浏览4
暂无评分
摘要
Recently, artificial intelligence has made remarkable advances and is being extensively utilized in many fields. However, there are challenges in adapting to real-world conditions in the field of face recognition where even minor errors are not tolerated. To address this, we propose a face recognition system that can remove brightness and contrast from images, which are common real-world conditions, without retraining the model. Our system can adapt to brightness and contrast accordingly using a brightness and contrast removal algorithm. The algorithm generates nearly identical images for comparison by removing brightness and contrast. Experimental results demonstrate high face recognition accuracy compared to existing systems, improving true positive and false negative accuracy by 4.55% and 10.34%, respectively, with a minimal computing overhead of only 1.40%. Overall, the face recognition accuracy improved by 4.60%.
更多
查看译文
关键词
Face Recognition,Brightness and Contrast,Face Dataset,Artificial Intelligence
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要