Scholarlink

In Vietnam, research and design activities related to recognition systems are making initial strides within universities. Previous scientific research and projects predominantly focused on areas like fingerprint recognition to support security systems. However, topics related to human-computer interaction through facial recognition have yet to be fully developed. This paper aims to elucidate various aspects of the SSD model, DeepFace, IrisNet, and FAISS, encompassing their operational principles and IoT applications in security technology through facial recognition for smart door locks. We perform to evaluate several famous models with existing algorithms. The results indicate that the system achieved nearly 93% accuracy for single faces with an execution time of 0.25 s for the Iris model.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Face Recognition for Big Data Using Search Engine for Smart System

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Phat Nguyen Huu,
  • Duong Nguyen Tung,
  • Khanh Nguyen Hoang Nam,
  • Quang Tran Minh

摘要

In Vietnam, research and design activities related to recognition systems are making initial strides within universities. Previous scientific research and projects predominantly focused on areas like fingerprint recognition to support security systems. However, topics related to human-computer interaction through facial recognition have yet to be fully developed. This paper aims to elucidate various aspects of the SSD model, DeepFace, IrisNet, and FAISS, encompassing their operational principles and IoT applications in security technology through facial recognition for smart door locks. We perform to evaluate several famous models with existing algorithms. The results indicate that the system achieved nearly 93% accuracy for single faces with an execution time of 0.25 s for the Iris model.