Traffic Signal classifier is an Artificial Intelligence model that can detect the traffic signals from the boards.

pip install -r requirements.txt

Running the Program!

python traffic.py

Boom! you got it working!

In load_data function i wasn't able to read the image, I tried os.read, and after sometime with mode='rb' i got the same error. then after a lot of search, i found that the image is in the form of a numpy array and i tried to read it using cv2.imread and finally it worked! then slowly I created that function and it worked! In get_model function, I was using gtsrb-small model from the start because I thought it would be easier to train. but then as I did it I didn't know where the error was comming from because I did everything right, applied a lot of filters, pooling, but still got a error with this name:

ValueError: Shapes (None, 3) and (None, 43) are incompatible

but then after a lot of time i found that the error was because of the shape of the image and the model. so i changed the model to the gtsrb model and it worked!