Research Article
A Hybrid WOA-MTCNN Algorithm for Accurate Face Detection Based on IoT Architecture
@INPROCEEDINGS{10.1007/978-3-031-70507-6_22, author={Xi Hu and Hang Ruan and Xin Xiong and Siqi Zhang and Ning Huang and Jun Wang}, title={A Hybrid WOA-MTCNN Algorithm for Accurate Face Detection Based on IoT Architecture}, proceedings={IoT as a Service. 9th EAI International Conference, IoTaaS 2023, Nanjing, China, October 27-29, 2023, Proceedings}, proceedings_a={IOTAAS}, year={2024}, month={10}, keywords={Convolutional Neural Network (CNN) Deep learning Face detection Image pyramid Whale Optimization Algorithm (WOA)}, doi={10.1007/978-3-031-70507-6_22} }- Xi Hu
Hang Ruan
Xin Xiong
Siqi Zhang
Ning Huang
Jun Wang
Year: 2024
A Hybrid WOA-MTCNN Algorithm for Accurate Face Detection Based on IoT Architecture
IOTAAS
Springer
DOI: 10.1007/978-3-031-70507-6_22
Abstract
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 batchsize of MTCNN to estimate the model. Secondly, we utilize the WOA algorithm to further optimize the weights and batchsize 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.
Copyright © 2023–2025 ICST
