Scholarlink

To further improve the detection accuracy in face detection based on IoT architecture, this paper designs and implements a hybrid face detection system based on the MTCNN model, and optimize the MTCNN model by combining Whale Optimization Algorithm (WOA). Firstly, we select and train the loss function weights and batch_size of MTCNN to estimate the model. Secondly, we utilize the WOA algorithm to further optimize the weights and batch_size parameters. Thirdly, we train and simulate the prediction by applying the obtained optimal parameters. Finally, we design and implement a face detection software based on the MTCNN model to test the practical application capabilities of the 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

A Hybrid WOA-MTCNN Algorithm for Accurate Face Detection Based on IoT Architecture

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Xi Hu,
  • Hang Ruan,
  • Xin Xiong,
  • Siqi Zhang,
  • Ning Huang,
  • Jun Wang

摘要

To further improve the detection accuracy in face detection based on IoT architecture, this paper designs and implements a hybrid face detection system based on the MTCNN model, and optimize the MTCNN model by combining Whale Optimization Algorithm (WOA). Firstly, we select and train the loss function weights and batch_size of MTCNN to estimate the model. Secondly, we utilize the WOA algorithm to further optimize the weights and batch_size parameters. Thirdly, we train and simulate the prediction by applying the obtained optimal parameters. Finally, we design and implement a face detection software based on the MTCNN model to test the practical application capabilities of the model.