This is pretty damn impressive - I'll definitely have to give it a deeper look later.
On seeing the name, I had hoped it had come out of Udacity and their Self-Driving Car Engineer Nanodegree program; I was a "beta-tester" for that course (read: in the first cohort of the new nanodegree), and the self-driving vehicle that was being developed as a part of that course had the same name.
As far as I know, the simulator and code and other stuff we used is all available on github:
Not as advanced, of course, but still another open-source avenue with a similar purpose. As I recall, it wasn't very difficult to work with, and even with a GTX 750 TI/SC GPU I was able to both render, train, and run models and still have 3D acceleration using CUDA on my Ubuntu install. Of course, that maxed out the GPU and put the fans on high, but it still worked fairly well at a decent resolution and level of detail settings. Just to show you don't need an expensive GPU to play around with this stuff...
The research is funded by a number of research centers, and yet there are advertisements for coca cola products littered throughout the simulator (I count 3 in the first 15 seconds). What gives? Is coke sponsoring this project? Why are they not listed as sponsors? Are we moving toward product placement advertisements in open source software?
Wow, they even let you simulate and control pedestrians? Between that and the pre-made, 3d, urban world to use as a testing ground...what a practical approach to getting these chops under your belt.
On seeing the name, I had hoped it had come out of Udacity and their Self-Driving Car Engineer Nanodegree program; I was a "beta-tester" for that course (read: in the first cohort of the new nanodegree), and the self-driving vehicle that was being developed as a part of that course had the same name.
As far as I know, the simulator and code and other stuff we used is all available on github:
https://github.com/udacity/self-driving-car
https://github.com/udacity/self-driving-car-sim
Not as advanced, of course, but still another open-source avenue with a similar purpose. As I recall, it wasn't very difficult to work with, and even with a GTX 750 TI/SC GPU I was able to both render, train, and run models and still have 3D acceleration using CUDA on my Ubuntu install. Of course, that maxed out the GPU and put the fans on high, but it still worked fairly well at a decent resolution and level of detail settings. Just to show you don't need an expensive GPU to play around with this stuff...