NVIDIA is a company known to make graphics card and computers since a long time ago. Most of its products are found in electronic products such as computers. Recently, the company has put in a lot of effort to go into autonomous car market by coming up with a new system which is able to do teach itself on how to handle scenarios which have not encountered before when driving on the public road with minimum interference from human drivers.
Most of autonomous cars which are currently being tested on public road able to drive based on a certain conditions present on the road. One example is it is able to stay in a line by detecting the painted lines on the highways. However, if the painted lines are not present, the system does not know what to do. Another challenge that an autonomous car face is driving on a snowy weather conditions where there is difficulty in differentiating the objects.
Apparently, NVIDIA has come up with a system which is able to handle those kinds of scenarios. The company has found an algorithm which is able to classify road even though there is no painted lines present on it as well as it is able to differentiate objects when driving through snowy weather.
The company explained how it works on the below video.