Udacity created a self-driving simulator for its self-driving car nanodegree so students could learn how to train cars to navigate road courses using deep learning. This project isn’t limited to the students of Udacity anymore; since Udacity open-sourced the simulator, anyone can use it.
The simulator lets any developer or hobbyist mess with Unity, the free game-making engine. Unity is necessary to load all of the assets for this simulator.
Since Udacity is building an open-source self-driving car, they want developers’ help, and this GitHub project is just one way to join the effort. Right now, the self-driving problem is broken down into multiple complex challenges, so developers and their team can compete to create a solution for a real self-driving car, according to Udacity.
Some of the challenges developers can start with include a self-driving car Android dashboard, which is about building a user-friendly, engaging dashboard for the Udacity car. This particular challenge asks developers to build something that lives up to expectation of stepping into a self-driving car, so it should be “futuristic, slick, performant, and above all: useful,” writes Oliver Cameron, the self-driving lead at Udacity.
Previous challenges include an image-based localization challenge, a 3D model for a camera mount challenge, and a deep learning steering angle challenge. The next challenge is coming soon, according to the Udacity self-driving page.
Udacity’s self-driving car engineering nanodegree lets anyone run code on an actual autonomous vehicle. They’ll also get their hands dirty with topics like deep learning, computer vision, sensor fusion, controllers, and automotive hardware.
#2. Lectures: Oxford’s Deep NLP course; contains slides in the repository.
#3. Wuzz: Interactive command-line interface tool for HTTP inspection.
#4. Lottie for iOS: Same as above, but for iOS.
#5. FreeCodeCamp: FreeCodeCamp is still on the list. Forever and ever trending.