An Intelligent System for Detecting People from Images and Videos Provided by IoT Devices

Today, thanks to artificial neural networks, the field of computer vision is undergoing an unprecedented revolution. The rise of deep learning and the advent of big data have both facilitated the construction of neural networks, not only deep, but also increasingly efficient. This has resulted in the automation of Internet of Things (IoT) technologies and devices. In addition, there are more and more disasters across the world where it is sometimes difficult to find victims in disaster areas. The contribution of IoT technologies is very important in this field. Thus, we propose in this work an intelligent system for detecting people from images and videos provided by IoT devices. Thus, we initially list some open source datasets used for object detection tasks. Then, we propose four person detection models based on the YOLO architectures. After evaluation, we obtained good performance results in terms of Average Accuracy.