MISSING PERSON IDENTIFICATION SYSTEM USING DEEP LEARNING

Paper ID: EIJTEM_2020_7_2_14-18

Author's Name: Dr. L. Sridhara Rao, A. Thanmai, M. Hemanth and N. Balraj

Volume: 7

Issue: 2

Year: 2020

Page No: 14-18

Abstract:

Every year, India receives innumerable reports of missing persons. A sizable portion of the missing people cases are still unsolved. This research describes an unique facial recognition approach for using deep learning to identify the reported missing people from the large number of child photographs that are accessible. Images of questionable people can be uploaded by the public onto a shared site along with annotations and locations. The image will be automatically compared to the repository's recorded images of the missing. The input image is classified, and the missing person's database photo with the best match is chosen. To do this, a deep learning model is trained to accurately distinguish the missing individuals from the missing individuals. Here, face identification is accomplished using the Convolutional Neural Network (CNN), a very successful deep learning technology for image-based applications. “With the aid of the pre-trained CNN model VGG Face deep architecture, face descriptors are retrieved from the pictures. In contrast to conventional deep learning applications, our technique simply employs the convolution network as a high level feature extractor, while the learned SVM classifier does the person recognition.”

Keywords: CNN, VGG face deep architecture, Support Vector Machine, Deep learning

View PDF