Autonomous (model sized) vehicles development with deep learning solutions

Abstract

This thesis explores the use of deep learning techniques for creating an autonomous car, specifically utilizing a camera and OpenCV for lane detection and steering angle prediction. While the study demonstrates the potential for DL in improving the accuracy and reliability of autonomous vehicles, several challenges and limitations must be addressed, including training process optimization, model robustness, and ethical and social implications. This work lays the foundation for future research in autonomous vehicle development and may hasten the introduction of dependable autonomous cars, altering transportation networks and mobility.