This article introduces a type of unmanned driving simulation environment-Udacity's unmanned driving simulation environment.
1. Udacity
Udacity is a "cloud university" founded by Sebastian Thrun, the father of Google's self-driving cars (and the author of the book "Probabilistic Robotics"). It gives everyone the opportunity to learn artificial intelligence, data science, and automation. Driving, natural language processing, computer vision, AI quantitative investment, blockchain, cloud computing, etc. For students who have completed their studies and passed the test, they can get a nanodegree issued by Udacity (I don't know the degree of recognition), which also introduces help to find a job, that is, employment support.
Take a look at the setting of driverless engineering courses:
They are all big cows!
The course is nearly RMB 8,000 , and the price/performance ratio may not be that high for me. But if you want to quickly enter a certain field in a short time (for example, driverless), and hope to find a good job. I think it's really not a loss to sign up for this course. I remember junior year, there was a roommate in order to find a java development-related work (I have electronic information engineering, and consequently learn some, are not proficient, shut away from the real-world engineering practice Cihao Duo ), spent sixteen thousand Thousand yuan applied for a java engineer training class.
There is a free trial lesson, and I listened to a less than one minute lesson on how to extract lane lines based on image processing. The course is indeed well-designed, and the theory and the code go hand in hand. I myself don’t rule out that when I don’t have much money in the future, I will choose a course to study hard (not necessarily driverless).
2. Udacity's driverless simulation environment
In the driverless engineer course of Udacity, in order to facilitate students to test the effectiveness of their own code, there will be some typical driverless simulation environment projects (path planning, end-to-end deep learning, etc.).
These simulation environments are packaged environments with specific scenarios and specific uses developed based on Unity. For example these two:
-
udacity/self-driving-car-sim
This is a project that uses deep learning-based unmanned vehicles. We can use it to train an end-to-end deep learning, or we can use our own training, or the project's own Model-driven unmanned vehicles in the simulation environment drive along the road and overtake other behaviors. The installation method is not difficult. The simulation environment is already packaged by others. Just download the executable file corresponding to your system (windows, Mac, Linux) (there is a download address). Then, git clone the project to the local, after the compilation is successful, run. Remember, only two things are needed: the packaged simulation environment and the corresponding git project.
-
udacity/CarND-Path-Planning-Project
This is a project about unmanned vehicle path planning. The configuration steps are the same as the above environment, first download the simulation environment corresponding to the system, and then git clone the project to the local, compile and run.
The following figure is a video of running the compiled code to control the operation of the car.
to sum up
The simulation environments introduced in this article have specific application purposes, and it is best to run with the corresponding project. Its advantage is that we can use the packaged simulation environment and project code to quickly and easily enter a knowledge gap field. Since the interface of the emulator is not very friendly, if you want the emulation environment to really serve you, you have to spend time to understand all the interfaces and data flows of the emulation environment. In addition, there are many other projects in the github/Udacity directory. I really feel that I have come to the sea of knowledge. I see if this is good and that is good. My eyes are full of light. It's just that the energy of a person is so much, I really have to study hard.