Learning to Look Around: Intelligently Exploring Unseen Environments for Unknown Tasks

TL;DR

This paper introduces a reinforcement learning approach for visual agents to autonomously explore and observe unseen environments, learning efficient behaviors for scene and shape completion without task-specific training.

Contribution

It presents a novel RL-based method enabling agents to learn general exploration policies that transfer across tasks and environments without manual labeling or recognition-specific training.

Findings

Policies effectively reduce uncertainty in unseen environments

Approach generalizes to new tasks and environments

No manual labeling required for training

Abstract

It is common to implicitly assume access to intelligently captured inputs (e.g., photos from a human photographer), yet autonomously capturing good observations is itself a major challenge. We address the problem of learning to look around: if a visual agent has the ability to voluntarily acquire new views to observe its environment, how can it learn efficient exploratory behaviors to acquire informative observations? We propose a reinforcement learning solution, where the agent is rewarded for actions that reduce its uncertainty about the unobserved portions of its environment. Based on this principle, we develop a recurrent neural network-based approach to perform active completion of panoramic natural scenes and 3D object shapes. Crucially, the learned policies are not tied to any recognition task nor to the particular semantic content seen during training. As a result, 1) the learned "look around" behavior is relevant even for new tasks in unseen environments, and 2) training data acquisition involves no manual labeling. Through tests in diverse settings, we demonstrate that our approach learns useful generic policies that transfer to new unseen tasks and environments. Completion episodes are shown at https://goo.gl/BgWX3W.