JetRacer Autonomous Driving

[]()

Description

Autonomous driving for Final Degree Project in Robotics.

Software

The employed architecture of convolutional neural network (CNN) is YOLOv8n-seg. This model has been trained with a dataset of 500 images of the following circuit (Figure 1). The achieved metrics of this model are:

Development

List TODO:

• [x] [Test JetRacer motors]()

• [x] [Keyboard Control JetRacer]()

• [x] [Gamepad Control JetRacer]()

• [x] [Create own dataset]().

• [x] [Train YOLO]().

• [x] [Lane segmentation]().

• [x] [Object segmentation]().

• [x] [Car follower]().

• [x] [Security stop]().

• [ ] [Implement EKF]().

License

JetRacer Autonomous Driving is available under next license:

• AGPL-3.0 License: See LICENSE file for details.

  Author:

  (c) 2023 (Marcos Fernández)