单选题
Google researchers have used artificial intelligence (AI) to teach robots how to move like real animals (in this case, dogs). They describe their experiment in a blog released this week. 'First, we describe how robots can learn to move like a real animal by imitating their motions, producing fast and fluent movements like running slowly and hopping. Then, we discuss a system for automating the training of moving skills in the real world, which allows robots to learn to walk on their own, with minimal human assistance,' shared in the blog Xue Bin (Jason) Peng, Student Researcher and Sehoon Ha, Research Scientist, Robotics at Google. They achieved this impressive feat by using something called reinforcement learning (RL). They began by taking a reference video recorded from an animal and using RL to get the robot to imitate an animal's movement. 'By providing the system with different reference motions, we are able to train an animal-shaped robot to perform a diverse set of agile behaviors, ranging from fast walking to dynamic hops and turns. The policies are trained primarily in simulation(模拟,模仿), and then transferred to the real world using a space adaptation technique that can efficiently adapt a policy using only a few minutes of data from the real robot,' wrote the researchers in their blog. However, it is a well-known fact that simulators provide a poor approximation of the real world, meaning that simulations don't perform well in reality. This is where the researchers decided to use a sample-efficient space adaptation technique. They did so by introducing an element of randomness to the physical parameters(参数) used in the simulation by varying physical quantities, such as the robot's mass and friction. This resulted in a machine learning model that could account for all kinds of small changes and the complications they create. The end result is a robot that moves like a real dog. This kind of work is crucial as it can open opportunities to use robots to do sophisticated tasks in the real world.
单选题
What is Google researchers' experiment about? ______