Traffic Sign Detection
Last updated:
0 purchases
Description:
This project aims to detect five types of traffic signs using transfer learning with two pre-trained models: RetinaNet and YOLOv5. The dataset is custom-built from dashcam footage collected on Vietnamese roads. It provides a comprehensive approach to detect traffic signs in real-world conditions and classifies signs not belonging to these five categories as "Other Signs."
Features:
- Custom Dataset: Built from dashcam footage with manually annotated labels using LabelImg.
- Detection Categories:
- No Entry
- Turn Right Only
- Yield
- Pedestrian Crossing
- No Parking
Requirements:
- Python 3.8+
- PyTorch
- OpenCV
- LabelImg for dataset annotation
- YOLOv5 repository
- RetinaNet repository
Instructions:
Install dependencies:
License:
For personal and professional use. You cannot resell or redistribute these repositories in their original state.
Files In This Product: (if this is empty don't purchase this product)
Customer Reviews
There are no reviews.
