Self-Driving Cars to Drive Autonomously using Deep Learning

The research ‘s primary objective is the development of a simulated model for driverless cars, with the eventual goal of integrating this model as software into real-world autonomous vehicles. This endeavor encompasses several key components, including the enhancement of image quality without compromising its integrity, improving the accuracy of turn detection, and enabling the autonomous vehicle to navigate on a generalized track with consistent training and validation accuracy. Additionally, the research focuses on fundamental capabilities such as mapping, tracking, and local planning, all of which are essential for autonomous driving. To address safety and collision avoidance, obstacle and curb detection methods are being implemented, as well as road vehicle tracking techniques to handle diverse traffic scenarios. The final goal is to create a robust autonomous vehicle capable of autonomously performing tasks such as lane changes, parking, and U-turns, effectively utilizing innovations in obstacle detection, curb detection, and vehicle tracking. These objectives collectively aim to propel autonomous driving technology towards practical real-world applications. However, it is crucial to consider the legal, ethical, and safety aspects in parallel with technological advancements when working on autonomous vehicles.