A deep neural network based toddler tracking system. (27th September 2021)

• Record Type: Journal Article
• Authors: Guney, Hanife; Aydin, Melek; Taskiran, Murat; Kahraman, Nihan
• Other Names: Wu Chase guestEditor.; Yildirim Tulay guestEditor.; Ivanovic Mirjana guestEditor.; Bellatreche Ladjel guestEditor.; Wyrzykowski Roman guestEditor.; Ciorba Florina M. guestEditor.
• Is Part Of: Concurrency and computation. Volume 34:Number 14(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-09-27
• Subjects: deep neural networks -- face recognition -- object detection -- object tracking -- person tracking -- toddler tracking system; Parallel processing (Electronic computers) -- Periodicals; Parallel computers -- Periodicals; 004.35
• DOI: 10.1002/cpe.6636
• Languages: English
• ISSNs: 1532-0626

Abstract

Summary: With the recent technological developments, auxiliary smart systems have started to be used in many issues that are important in people's lives. One of these issues is that parents ensure the safety of their toddlers. Toddlers tend to reach objects that may be dangerous to them at home because of their natural curiosity. For this reason, parents' inability to pay enough attention to their toddlers while doing their daily work can cause irreversible accidents. On account of all these situations, it is clear that there is a need to develop a toddler tracking system. In this article, a toddler tracking system based on deep neural network has been proposed. For experimental studies, videos in 10 different scenarios were collected from different toddlers aged between 1.5 and 4. The system includes FaceNet for face recognition and YOLOv3 for object tracking. First, desired toddler's face is recognized, and the system tracks the movements of that toddler. If the toddler gets close to dangerous objects in the house, the parents will be warned by the system. As a result of experimental studies, the system accuracy rate was increased from 80.7% to 93.3% with the changes in the previous system.