Hands on Labs

Traffic Sign Recognition Classifier with TensorFlow

Room 32, 3rd floor

21st November, 15:00-16:00

During the "hands on" lab we will use notions about deep neural networks and convolutional neural networks to classify traffic signs. We will train and validate a model so it can classify traffic signs based on the "German Traffic Sign Dataset". With the trained model we will then try to predict the class of German traffic sign images found randomly on the web.

The goals of this project are the following:

  • Load the data set
  • Explore, summarize and visualize the data set
  • Design, train and test the model architecture
  • Use the model to make predictions on new images
All this while getting a glimpse on how to use Tensorflow from Python.

Adrian Fatol


Is a declared Machine Learning enthusiast. He finds passion in self-education about deep learning and studying about self-driving cars. All the knowledge he gathers, transforms in expertise in the travel industry, where he finds his favourite “playground” by working on various projects that imply scoring hotel image quality or record linkage for finding duplicate items.
On the other side of the playground, is the "serious" job as a Software Architect @Arobs Transilvania applying technologies like Python, Golang, Nodejs, Php, Mysql, Postgresql, Mongodb, Javascript and everything web related. He works especially on solving scalability, availability, performance and security problems in stacks, such as Kubernetes, Docker, Linux, Microservices.
Already a keynote speaker at four editions Codecamp that took place in Cluj , Baia Mare and Timisoara, he’s more than open to gladly give a glimpse of his knowledge.
In the last years, he was present as a speaker at Codecamp in Cluj, Baia Mare & Timisoara. He’s glad pay forward his knowledge to the IT community and always trying to motivate others to learn new things.