Udacity's Self-Driving Car Nanodegree team and its instructors write about the curriculum, autonomous vehicle careers, and research reviews.
HN links it for self-driving coverage; commenters discuss the course projects, safety statistics, and adversarial traffic signs.
I think the SDCND is a great opportunity to get hands on experience with cutting edge technologies. You will be evaluated and receive personalized feedback on about 15 practical assignments covering these topics:
Computer vision (Detect lane lines in a variety of conditions, including changing road surfaces, curved roads, and variable lighting), Neural networks (classify traffic signs, drive a car in a simulator), Track vehicles in camera images using image classifiers such as SVMs, decision trees, HOG, and DNNs.
More projects will cover these topics: Sensor Fusion, Localization, Control, Path Planning, Systems and an Elective.
If you know of comparable ressource to gain experience with all this material please share, I'm not aware of any.
Official source: http://medium.com/self-driving-cars/term-1-in-depth-on-udacitys-self-driving-car-curriculum-ffcf46af0c08
That seems to be an article about how David Silver intends to write some articles on the subject.
Do you know how to find the actual article on the subject?
*edit:
Found it! Google is a much better way to search medium than medium's worthless navigation/search.
Here are the three articles:
https://medium.com/self-driving-cars/how-to-land-an-autonomous-vehicle-job-coursework-e7acc2bfe740#.a343s2fsu
https://medium.com/self-driving-cars/how-to-land-an-autonomous-vehicle-job-projects-60e073a29bab#.l0j12xs85
https://medium.com/self-driving-cars/how-to-land-an-autonomous-vehicle-job-networking-d2bcd6a8e656#.g7guh92jc
Next decade, the NYT will run another story titled: "Biggest drop in traffic deaths in 60 years".
Autonomous cars are around the corner and it's the only way to truly make a leap forward in terms of traffic safety.
And people can spend even more time on their apps while commuting, uninterrupted by this soon-to-be-forgotten distraction called "manual driving".
Incidentally wrote a blog about this today: https://medium.com/self-driving-cars/safer-roads-and-cleaner-air-with-autonomous-fleets-91b5d879be68#.1m9yo6lm7
Most interesting to think about are the stickers[1] that look random but actually trigger some kind of adversarial behaviour, a bit like how tiny human-imperceptible changes to an image can completely flip an image categorisation network's guess.
Perhaps most fun to think about in a Dan Brown-style NSA capabilities combined with Peter Watts' Blindsight's near-total-imformation perceptual-cognitive shenanigans kind of way.
[1]: https://medium.com/self-driving-cars/adversarial-traffic-signs-fd16b7171906
That video isn't an example of this. Tesla's statements about fleet learning are pretty much theoretical - they're talking about a system they want to build someday. Right now, their motion planning and prediction is all hardcoded robotics.
See Karpathy's talk here: https://medium.com/self-driving-cars/annotated-karpathys-autopilot-talk-3e6270a21f8d
Not sure it's that clear. Using data from Wikipedia, I got 11 deaths per billion km. The 4 million miles are just from Weymo, not all autonomous cars in total. (6 million+ https://medium.com/self-driving-cars/miles-driven-677bda21b0f7) And the confidence level on the 1 event over 6+ million miles is not great either.