This project implements rider helmet detection using YOLOv7 (an object detection model) and DeepSORT (an object tracking model) on a cctv recording. This is a group effort where each group were assigned to train combination of different hyperparameter setting and video time range and manually label images from the set of video frames using roboflow.

link to google colab