AlessandroFerrante/StreetSignSet

Street Sign Set

---

---

High-Quality Traffic Sign Detection Dataset

📂 Dataset Overview

Street Sign Set is a comprehensive dataset designed for road sign detection in realistic contexts. It serves as the foundation for the StreetSignSense project, enabling robust detection in diverse environmental conditions.

The dataset is not perfectly balanced, reflecting the real-world frequency where some signs appear much more often than others.

📊 Dataset Statistics

• Total Images: > 7,300 images.
• Classes: 63 distinct classes.
• Macro-Categories: 5 (Priority, Prohibition, Information, Warning, Mandatory).
• Format: Standard YOLO annotations (.txt).

🏷️ Class Structure and Labels

The 63 classes are organized into 5 macro-categories that define the label prefix:

1. prio (Priority) - e.g., prio_give_way, stop
2. forb (Prohibition) - e.g., forb_speed_over_50
3. info (Information) - e.g., info_parking
4. warn (Warning) - e.g., warn_right_curve
5. mand (Mandatory) - e.g., mand_pass_left_right

Primary Targets (23 Main Classes)

The dataset focuses on 23 main classes identified as primary targets, including:

• Speed limits: 14 classes (e.g., 5–130 km/h).
• Prohibition signs: 4 classes (e.g., no stopping/parking, no overtaking).
• Priority signs: 2 classes (e.g., give way, stop).
• Curves and crossings: 3 classes (e.g., dangerous curves, pedestrian crossing).

🛠️ Hybrid Origin and Construction

This dataset is a result of a hybrid curation process:

• Base: ~4000 images sourced from existing Kaggle datasets.
• Expansion: ~3000 images manually integrated from external sources and street mapping services to cover underrepresented classes. These were manually labeled to ensure quality.

⚙️ Technical Specifications

• Filename Scheme: Rigorous logical scheme class_name-n.jpg (e.g., prio_give_way-12.jpg).
• Selective Data Augmentation: Applied only to rare classes to mitigate class imbalance. Techniques include:
  • Hue/Saturation/Brightness variations.
  • Grayscale (23% probability).
  • Blur and Noise simulation for adverse conditions.

📥 Download & Access

To keep the GitHub repository lightweight, the raw dataset is hosted on external platforms specialized for data versioning.

🖊️ Citation

If you use this dataset in your research, please cite it as follows:

@misc{alessandro_ferrante_2025,
    title={Street Sign Set},
    url={[https://www.kaggle.com/ds/8410752](https://www.kaggle.com/ds/8410752)},
    DOI={10.34740/KAGGLE/DS/8410752},
    publisher={Kaggle},
    author={Alessandro Ferrante},
    year={2025}
}

Dataset Structure

The data is organized following the standard YOLO convention, making it ready for immediate training:

.
├── train/
│ ├── images/ # Training set
│ └── labels/ # YOLO annotations
├── val/
│ ├── images/ # Validation set
│ └── labels/ # YOLO annotations
├── test/
│ ├── images/ # Test set for final evaluation
│ └── labels/ # YOLO annotations
├── data.yaml # Dataset configuration file (classes names)
└── dataset_analysis.csv # Detailed analysis of the dataset class distribution

👨‍💻 Author

Alessandro Ferrante

Email: [email protected]