Teaching robots new skills is going to be much easier in the future. Openai has already developed a robotic system trained in simulation that allows physical robots to learn an new task after seeing it done once. The one-shot imitation learning algorithm allows humans to show a robot how to do a task by performing it in VR,
The system is powered by two neural networks for vision and imitation. Here is how it works:
The vision network ingests an image from the robot’s camera and outputs state representing the positions of the objects. The imitation network observes a demonstration, processes it to infer the intent of the task, and then accomplishes the intent starting from another starting configuration. Thus, the imitation network must generalize the demonstration to a new setting. The network learns this from the distribution of training examples. It is trained on dozens of different tasks with thousands of demonstrations for each task.
Pretty fascinating, don’t you think? more info is available here.
