Seat Belt Detection

Project Overview

This project involves creating a machine learning model to detect the presence of seat belts in vehicle images. The primary goal is to enhance road safety by automating the detection process, which can be used in surveillance systems, automotive safety checks, and more.

Descriptions of Each Class Type

• Seat Belt Worn: Images where the seat belt is correctly fastened.
• Seat Belt Not Worn: Images where the seat belt is not fastened.
• Not Clear: Images where it's not possible to determine if the seat belt is fastened due to obstructions or poor image quality.

Current Status and Timeline

• Data Collection: Completed.
• Annotation: Ongoing.
• Model Training: In progress.
• Evaluation: Planned.
• Deployment: Expected in the near future.

Contribution and Labeling Guidelines

• Contributions: Community contributions are welcome. Please fork the repository and submit a pull request with detailed descriptions of your changes.
• Labeling Guidelines:
  • Ensure images are clear and the seat belt is visible.
  • Use high-quality images in various scenarios (day/night, different vehicle types, etc.).
  • Label images consistently according to the defined classes.

For any questions or issues, please contact karanpanja123@gmail.com

                        @misc{
                            seat-belt-detection-2_dataset,
                            title = { seat belt detection 2 Dataset },
                            type = { Open Source Dataset },
                            author = { SeatBeltMobileUsageDetection },
                            howpublished = { \url{ https://universe.roboflow.com/seatbeltmobileusagedetection/seat-belt-detection-2 } },
                            url = { https://universe.roboflow.com/seatbeltmobileusagedetection/seat-belt-detection-2 },
                            journal = { Roboflow Universe },
                            publisher = { Roboflow },
                            year = { 2024 },
                            month = { aug },
                            note = { visited on 2024-12-18 },
                            }