DEEP LEARNING IMPLEMENTATION IN AUTONOMOUS DRIVING CAR SIMULATION BY USING AIRSIM

• Main Author: Gunova, Violla
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/56942
• Institution: Institut Teknologi Bandung

Machine Learning is one of the aspects in Artificial Intelligence where this aspect makes machines to be able to learn and become smarter like humans. There are various current technologies that apply Machine Learning, one of them is Autonomous Driving technology. This technology aims to help human work related to driving. There are several methods in Machine Learning that can be used to build this technology. In order to build this technology without taking a long training-time, Deep Learning and Supervised Learning methods will be used. The output of the establishment of Autonomous Driving technology this time is the car moves according to the trajectory and has short training-time. In order to achieve it, the training will be done by using 1-set of training data which is recorded in order to find the best training parameters at first. After the parameters are obtained, the training will be executed with various-sets of training data using these parameters. The number of set data for training will be added continuously until the initial outputs of establishing this technology are achieved. This technology are done by using the help of AirSim simulator. This must be done to avoid the damage that occurs when the exercise is done in the real world.