Keywords: Computer vision, reinforcement learning, unsupervised learning, active learning
TL;DR: An active-vision agent can learn to move a camera to informative points of view without external reward, by using prediction errors of its own proprioceptive sensors as a negative reward.
Abstract: We present a method for learning active vision skills, for moving the camera to observe a robot's sensors from informative points of view, without external rewards or labels. We do this by jointly training a visual predictor network, which predicts future returns of the sensors using pixels, and a camera control agent, which we reward using the negative error of the predictor. The agent thus moves the camera to points of view that are most predictive for a target sensor, which we select using a conditioning input to the agent. We show that despite this noisy learned reward function, the learned policies are competent, avoid occlusions, and precisely frame the sensor to a specific location in the view, which we call an emergent fovea. We find that replacing the conventional camera with a foveal camera further increases the policies' precision.