This project applies machine learning techniques to a custom dataset focused on bus pad detection. The goal is to develop a model that accurately identifies and classifies bus pads in urban environments using deep learning frameworks.

• Custom dataset tailored for bus pad detection
• Data preprocessing and augmentation techniques
• Model training using YOLOv5
• Evaluation and validation of model performance
• Deployment-ready inference script

The dataset consists of annotated images focusing on bus pads in various urban conditions. Each image is labeled in the YOLO format (class x_center y_center width height) to assist in model training and evaluation.

dataset/ ├── images/ │ ├── train/ │ ├── val/ │ ├── test/ ├── labels/ │ ├── train/ │ ├── val/ │ ├── test/ 

The dataset configuration file (dataset.yaml) must be updated as follows:

train: path/to/train/images val: path/to/val/images test: path/to/test/images nc: <number_of_classes> names: ['class1', 'class2', ...]

1. Clone the repository:

   git clone https://github.com/your-repo/machine-learning-buspads.git

2. Install dependencies:

   pip install -r requirements.txt

3. Prepare the dataset and place it in the data/ directory.

Run the following command to start training:

python train.py --img 640 --batch 16 --epochs 50 --data dataset.yaml --weights yolov5s.pt --device 0

• --img: Input image size
• --batch: Batch size
• --epochs: Number of training epochs
• --data: Path to dataset configuration
• --weights: Pre-trained model weights (use yolov5s.pt, yolov5m.pt, etc.)
• --device: GPU/CPU selection (0 for GPU, cpu for CPU)

To evaluate model performance:

python val.py --data dataset.yaml --weights runs/train/exp/weights/best.pt --img 640

To test the model on new images:

python detect.py --weights runs/train/exp/weights/best.pt --img 640 --source path/to/images

The model's performance is assessed using accuracy, precision, recall, and F1-score. The following table summarizes the results:

• Increase dataset size for better generalization
• Improve annotation quality
• Optimize hyperparameters for better performance
• Implement real-time detection for deployment

This project is licensed under the MIT License.

For any inquiries, please contact your-email@example.com.

This project demonstrates how to train a YOLOv5 model using a custom dataset (Buspads). Further improvements can be made by optimizing hyperparameters and augmenting the dataset for better accuracy.

For any issues or contributions, feel free to open a discussion or pull request!