The Motorcycle Classification project aims to detect and classify motorcycles in images into five categories: Honda, Suzuki, Yamaha, VinFast, and Others. The project uses the YOLOv8n model by Ultralytics for motorcycle detection and classification. This is my Final Project in Mechina Learning(CS114) at university

Data in this project consists of 33,646 photos of motorbikes divided into 5 labels(Honda, Suzuki, Yamaha, VinFast, and Others) contributed by students in the class.

The project consists of the following steps:

1. Motorcycle Detection in Images:

   • The YOLOv8n model is used to detect motorcycles in images.
   • The detection process outputs the bounding boxes containing the motorcycles, along with their positions and dimensions.

2. Create COCO-formatted Data Files:

   • The detection results are stored in COCO format, with the following fields:
     • label: The motorcycle category (0: Others, 1: Honda, 2: Suzuki, 3: Yamaha, 4: VinFast).
     • center_x: The x-coordinate of the bounding box center.
     • center_y: The y-coordinate of the bounding box center.
     • bbox_width: The width of the bounding box.
     • bbox_height: The height of the bounding box.

3. Train the Model:

   • The YOLOv8n model is trained on the prepared dataset in COCO format.
   • The training process includes data preprocessing, model training, and hyperparameter tuning.

4. Evaluate the Model:

   • After training, the model is evaluated on a validation dataset to measure accuracy, recall, and other performance metrics.
   • The evaluation results are used to improve the model.

• Python 3.x
• Ultralytics YOLOv8