US specialists have created a special neural algorithm that allows robots to analyze the trajectory of a ball and more.
Robots have been taught to play table tennis — now the iron things can hit back over a hundred times in a row and even resist humans.
Specialists from the University of California at Berkeley have created a special neural algorithm that allows machines to analyze the trajectory of a ball and calculate the actions of their entire body at once.
Humanoids can already be used to train beginner tennis players, and in the future they want to pump them up to the level of professionals