🚗 PoliDuckie Reinforcement Learning 🚙

AIDO 2021 finalists!

Folder with our experiments on autonomous driving with Reinforcement Learning and the Duckietown platform.

Preprocessing

We tried line extraction and bird eye view. During training we saw improvements using line extraction while birdeye didn't improve our model.

Results

Team

• Ksenia Chigibaeva
• Giorgio Miani
• Toma Sikora
• Giulio Vaccari
• Davide Viviani
• Gabriel Voss