Its interesting, that kind of object level fusion is a fairly different problem to training visual perception, following some of the less in fashion robotics te

Its interesting, that kind of object level fusion is a fairly different problem to training visual perception, following some of the less in fashion robotics techniques. I wonder if its a case of the Tesla engineers focusing on the fad technologies (or just their strengths) more than its a hardware cost thing.