This project uses deep learning to detect and recognize traffic signs in a road image.

The classifer is trainded on 80 different classes.

Two technologies are used :

• YOLO "you only look once" to detect traffic signs.
• Convolutional neural network to classify traffic signs.

The code is implemented in python3 using Pytorch and Opencv. The web interface implemented using Stramlit.

To run the web interface:

1- install all the requirements

pip install -r requirements.txt

2- download the Yolo weights from Google drive

3- put all files in the same directory

4- to run the web interface

streamlit run web.py

• Karim khadro