Data Analytics
Moroccan Traffic Sign Detection
A computer-vision model that detects and recognises Moroccan traffic signs, trained on a dataset the team collected and annotated.
- Type
- Master programme project, ENSAK
- Role
- Data scientist
- Period
- Jan 2024 to Jan 2024
- Status
- Completed
285
collected and annotated images
YOLOv8
detector compared with a CNN baseline
21
days from start to results
The idea
Public models are trained on European and US signs. Moroccan roads have their own signs and conditions, and there was no ready dataset.
What I built
A dataset of 285 collected images annotated in Annotation Lab, a CNN baseline in TensorFlow and a YOLOv8 detector, compared on the same data.
The result
A trained detector and a published Kaggle dataset that others can reuse.
Key features
- Collected Moroccan dataset on Kaggle
- CNN baseline and YOLOv8 detector
- Training and evaluation notebook
Screens
exploring first dataset 1 / 5
Stack and tools
- YOLO
- Python
- TensorFlow
- Data
- Kaggel
- YOLO V8
- Annotation Lab
Data Sources
collected Moroccan traffic sign detection
285
Traffic Sign Preprocessed
4180
Code Samples
Traffic sign detection sript
LoadingNeed something like this built?
Book a free 30-minute call to talk through your project. After the discovery call, once the scope is clear, you get a fixed written quote in MAD.
