Role of Artificial Intelligence Techniques in Active Safety using Image Processing for Autonomous Driving Vehicles

Abstract

This paper presents a study on the importance of using Artificial Intelligence methods in developing self-driving vehicles. Advances in Artificial Intelligence are one of the key enablers of the Autonomous Vehicles development. There are several ways to increase the level of autonomy of a vehicle and make it capable to avoid or prevent crashes. Advanced Driver Assistance Systems will help autonomous vehicles become a reality. Image processing of various traffic scenarios can be drastically improved by the use of the superior degrees of computer processing and computer vision techniques. With the help of convolutional neural networks (CNNs), not only that a single object can be detected and tracked in a sequence, but all relevant objects can be detected and classified for further processing. CNN s, the current state-of-the art for efficiently implementing deep neural networks for vision, are more efficient because they reuse a lot of weights across the image. Also, an introduction in synthetic data generation field is presented as a way to overcome the lack of labeled datasets for training networks.