Bike Rider Helmet Detection using YOLOv8

This repository contains a walk through of implementation of bike rider helmet detection using YOLOv8. As we know that bike riders who do not wear helmet may which result in fatal accidents and death in some cases. The goal is to create a ML/DL model that can detect if a person is wearing helment or not. May sure to select GPU for running through the notebook.

Methodology :-

1. Data Collection:- Collect images/videos of bike rider, helmet and no helmet for the model to train upon.

2. Data pre-processing:- Pre-processing and autolabel images and videos using foundation model like DINO and SAM (Segment anything model).

3. Train YOLOv8 Model

4. Evaluate Target Model

5. Run Inference on images and videos

This repository output may change in the near future:-

https://drive.google.com/drive/folders/1M4FckJJeyPQTTWqgo6xWhW8L4tf0EJ4l?usp=sharing

Also see

1. https://github.com/Viddesh1/Bike-Helmet-Detection
2. https://github.com/Viddesh1/Bike-Helmet-Detectionv2
3. https://github.com/Viddesh1/Bike-Helmet-Detection-Docs