Helmet Detection Using Detectron2 and EfficientDet

In the past few years, there has been great technological advancement in the field of deep learning and its application in image classification, image segmentation and object detection. One of those applications is detecting helmets on bike riders. Since road accidents are increasing day by day, it is important that riders wear helmets all the time so that injuries can be avoided. In the past few years, many deep learning approaches have been made to solve this problem of detection of helmets on bike riders. A lot of research has been conducted in this field using primitive technologies like CNNs. But the recent advancement in computer vision techniques has shown a lot of promise. This paper acts as an analytical study of various different approaches and also experiments the new state of the art object detection models such as Detectron2 and EfficientDet. The results from these state of the art models show a lot of promise.