Self-Driving Car Module- A CNN Based Transfer Learning Model

Abstract

Humans' lives will be made easier by the commercial use of self-driving cars. This essay's goal is to discuss the subject. The most significant features of self-driving car technology are covered. Design examines the fundamental components of a self-driving car. The four primary technologies of a self-driving car are addressed and evaluated: a navigation system, path planning, environment perception, and automobile control. The research's findings, significant scientific advances, research successes, and the research institution have all been condensed. This study uses a variety of words, including path planning, vehicle control, vehicle navigation, environment perception, and self-driving automobiles. Vision-based navigation systems can now be considered as a possible substitute to traditional navigation sensors like Global Navigation Satellite Systems (GNSS) and Inertial Navigation Systems thanks to improvements in computer vision . Tesla, BMW, Mclaren, and TATA are all manufacturing electric cars with self-driving features, but this feature is only applicable for first world countries. Autonomous cars can not detect paths on Highway and intercity ways of second and third world countries, so transfer learning will help us to work on that. Self-driving cars will change how people commute, work, and play while also promoting a cleaner, safer environment. A safer environment would come from the employment of autonomous automobiles on the road as a consequence of prompt and wise decision-making, which would stop avoidable traffic congestion and fatalities from happening. As there are no record of same type research of work previously we can not compare the result but we got 94% of accuracy using this model.