This repository implements the OpenAI Gym interface to be compatible with AWS DeepRacer
- python 3.6
- tensorflow=1.15.x
For object detection we used this implementation of YOLO V3 deep neural network. To use the model you will need to download weights and convert them by runnig the following:
wget https://pjreddie.com/media/files/yolov3.weights
python convert.py yolov3.cfg yolov3.weights model_data/yolo.h5
Visit qqwweee/keras-yolo3 for mor instructions.
As for now, we do not automatically retrieve the cookie and x-csrf-token. In order to do that, follow the instructions here to connect your DeepRacer to the wifi. After the car is connected, surf to its console, capture the network traffic and extract the fields from one of the headers' request.