Design a Hybrid Neural Network Tracking System Using Multiple Cameras
Ahmad Ismail, Sergey V. Vishnyakov, Mohammad Zedan · 2024
Neural networks have become a transformative technology, enabling significant advances over the past few decades. In particular, detection and tracking systems benefit from the ability of neural networks to learn and extract features, resulting in accurate object classification and localization. In this paper, we propose a hybrid neural network framework that integrates data from multiple cameras positioned along an object's path. This multi-camera configuration expands the system's field of view and improves prediction accuracy when tracking an object's position. We conduct extensive experiments to evaluate the performance of our proposed model, which demonstrates superior tracking accuracy and robustness across different environments compared to traditional single-camera systems.