In the course 'More examples of what machine learning can and cannot do'

Andrew mentions that it is hard for the ML to recognize human gestures.

He mentions that 1. It is hard to collect data, and 2. It requires high accuracy. 2 I get it, but why 1? He mentions that even with 10,000 data sets its hard for ML to recognize hand gestures, intensions. Can someone explain why?

1 Like

Hello @cernos88
The reason why Andrew mentions that it is hard to collect data to recognize hand gestures people are doing at your car is, there is no standard format of hand gestures that people doing at your car either to slow down or to stop or to convey some other information to you.
The number of ways that people could gesture at you is just very, very large.

1 Like