Child Re-identification

Role in Project

• Researched and evaluated various architectures for person re-identification tasks. Selected and fine-tuned Omni-Scale Neural Networks, using child-specific datasets.
• Performed data collection and annotation for child ReID task, capturing real-world scenarios of children engaged in play at Robokids.
• Optimized ReID system to achieve +5FPS for 24 HD cameras at single GPU

Product Description

We have developed a real-time child re-identification system for the Robokids, aiming to enhance parental oversight, improve children’s play experiences, and provide more accurate user data.

Real-Time Parental Monitoring: This project enables parents to monitor their children in real-time within Robokids. The entire facility is equipped with surveillance cameras that are integrated with our child ReID system. Parents can remotely track their child’s activities through their mobile phones, giving them peace of mind while they attend to personal matters outside Robokids, such as shopping or watching a movie.

Accurate User Data Access: This project provides more precise and comprehensive user data. This data includes information such as the number of users, their age, gender, and the duration of their play sessions. With this system in place, the center will be able to track user behavior more effectively, which is crucial for improving customer experience, optimizing marketing strategies, and enhancing product quality.