Publication

Abstract : An essential component of many user applications today, as well as security surveillance systems, text recognition, and some diagnosis for diseases from few scans like CT or MRI, is object identification, a computer vision problem. Also if we can see one of the most important components like object detection which can help detect any object etc. The traffic sign recognition is also a main important task which have a crucial role in intelligent autonomous vehicles. In our work we have implemented different models of YOLO v5 (You only look once) algorithm for traffic sign detection. To train yolo algorithm for the traffic sign detection on German Traffic Sign detection benchmark which have multiple scenario of images like moving vehicle images, clear images, uncleared images, and different environmental conditional images all these will help in to detect the four different classes and improve the mAP of the model with the existing work and comparing results of different models of Yolo v5 algorithm with the existing work.