1. Introduction
Robotics approach and a deep learning approach
Robotics approach uses output from a suite of sensors to directly measure the vehicle surroundings and then navigate accordingly self-driving car engineers.
Deep neural networks allow self-driving cars to learn how to drive by mimicking human driving behavior.
Both the robotics and deep learning methods are actively being pursued today in the development of self-driving cars.
TERM 1
COMPUTER VISION AND DEEP LEARNING
PROJECT 1
FINGDING LANE LINES
A very first project in the self-driving candle will be to find lane markings. It is important.
If you can't find the markings you will have no clue where to drive.
PROJECT 2
BEHAVIORAL CLONING
We are here at rolling me instead our learning problem with copy and cologne you would behave yourselves are steering actions and brake gas actions.
And you get a chance to try it on your own when we train people who don't train them even the rules between environmental examples, so this unit you want to apply for that works deep learning to Cameron ridges and do some the very things in the surface are to the right.
PROJECT 3
ADVANCED LANE FINGDING AND VEHICLE DETECTION
What is computer vision but you and i drive a car we use eyes more than any other ogan understand what to do. Computer vision has the same capability into computers use cameras and we learn how to extract things like markings and other vehicles lot of cameras. It is fun.
TERM 2
SENSOR FUSION, LOCALIZATION AND CONTROL
PROJECT 4
SENSOR FUSION
What is sensor fusion? It is the science how to integrate different types of sensors