DIY Autonomous Cars 101: Recognizing Street Signs This is part of my article series. The project involves deep learninng to classify street signs. Full link to the blog post can be found here,
This project requires Python 2.7 with the following library installed:
The dataset used here is from Gernman Traffic Sign Benchmark website http://benchmark.ini.rub.de/?section=home&subsection=news
python attempt2.py should run the program properly, but if it won't it means you haven't set your directories properly.
Your directories should be in the following fashion.
Path | |attempt2.py |utils.py |vgg16.py | |GTSRB ---> |train |valid |results |sample (Optional)
Train folder will be the one containing the images with all classes, valid folder will be filled by the code.