Autonomous cars cannot drive. Here is why. (Ep. 210)
Data Science at Home
00:00
Using Deep Learning to Drive a Car
The idea of many engineers is like okay if there is temporal dependency and you know how can we leverage that. Instead of using for example common filter one can use GRUs a gated recurrent units which are kind of the backbone of long short term memory network at LSTM. If you rely heavily on deep learning or solely on deep learning I don't want to be in that car honestly, he says.
Transcript
Play full episode
