Skip to content

mairob/Trafficsign-classification-in-TensorFlow

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Practical introduction to convolutional neural networks (CNNs) in TensorFlow

Hi guys! My name Robert. I'm currently a student at OTH Regensburg. Due to a 2 months project work I had my first hands on experience with python, TensorFlow and convolutional neural networks. The jupyter notebook and the code that goes with it are some results of that work. All of that content is supposed to help sutdents getting started with TensorFlow and give some insights into CNNs.

What are we going to do?

<> Build a CNN for traffic sign classification with ~ 99.1 % accuracy based on the GTSRB <> Use TensorFlow and Python to do all the work <> Get a feeling on how CNNs work <> What to keep in mind when working with a CNN

Requirements

<> TensorFlow with GPU support (and a supported GPU!) <> 16 GB RAM (since we will do all the stuff quick and dirty ;) <> skimage and OpenCV for python 3.5.x

Note

The .ipynb-notebook has some images and gifs included which sum up to a total of ~18MB