Data Science and Intelligent Computing Techniques
Design and Implementation of IOT Based Face Detection and Recognition
Authors: Vijay Gaikwad, Devanshu Rathi, Vansh Rahangdale, Rahul Pandita, Kashish Rahate and Rajendra Singh Rajpurohit
Publishing Date: 09-02-2024
ISBN: 978-81-955020-2-8
Abstract
In contemporary society, the demand for accurate and efficient face detection in images and videos has grown significantly, driven by applications in surveillance, education, autonomous driving, and healthcare. Challenges such as unconstrained pose variations, occlusions, a large number of faces, and varying illumination conditions have posed formidable hurdles for existing face detection methods. In response, this study introduces a novel Depth wise Separable Convolution Block (DSCB) that not only maintains training speed but also enhances accuracy. Leveraging the proposed DSCB, a face detection model based on Multi-task Convolution Neural Network (MTCNN) is designed to tackle challenges related to occlusion, unconstrained pose variations, and numerous small targets. Compared to the original MTCNN, the proposed face detection method showcases substantial performance improvements and achievements. Furthermore, the paper delves into the realm of face recognition, an essential biometric technology that identifies individuals based on facial features. Traditional face recognition methods were plagued by slow processing speeds and lower accuracy compared to manual recognition. However, the advent of deep learning and Convolutional Neural Networks (CNNs) has revolutionized face recognition. Since the impact of pandemic moving towards physical interaction free life has become a norm. Also, with advancement in technology, recognition of individual via face has also spread its root in various daily use aspects of life like mobile phones, security safe, etc. This project aims to advance the home security systems via use of facial recognition. By addressing the associated challenges and concerns, the project aims to contribute to the ongoing development and adoption of smart doorbell IoT systems, ultimately leading to safer and more connected homes.
Keywords
MTCNN, Face Recognition, face Detection, Smart Doorbell, IoT.
Cite as
Vijay Gaikwad, Devanshu Rathi, Vansh Rahangdale, Rahul Pandita, Kashish Rahate and Rajendra Singh Rajpurohit, "Design and Implementation of IOT Based Face Detection and Recognition", In: Satyasai Jagannath Nanda and Rajendra Prasad Yadav (eds), Data Science and Intelligent Computing Techniques, SCRS, India, 2024, pp. 923-933. https://doi.org/10.56155/978-81-955020-2-8-78