Deep Active Localization
This work addresses the challenge of efficient robot pose disambiguation, offering a simulation-to-real transfer solution that is incremental in combining deep learning modules with existing frameworks.
The paper tackles the problem of active localization for robots by proposing an end-to-end differentiable method that learns to take informative actions, trainable in simulation and transferable to real hardware without refinement. The result shows that this system outperforms traditional approaches in perception or planning, with demonstrated robustness to various map configurations and nuisance parameters.
Active localization is the problem of generating robot actions that allow it to maximally disambiguate its pose within a reference map. Traditional approaches to this use an information-theoretic criterion for action selection and hand-crafted perceptual models. In this work we propose an end-to-end differentiable method for learning to take informative actions that is trainable entirely in simulation and then transferable to real robot hardware with zero refinement. The system is composed of two modules: a convolutional neural network for perception, and a deep reinforcement learned planning module. We introduce a multi-scale approach to the learned perceptual model since the accuracy needed to perform action selection with reinforcement learning is much less than the accuracy needed for robot control. We demonstrate that the resulting system outperforms using the traditional approach for either perception or planning. We also demonstrate our approaches robustness to different map configurations and other nuisance parameters through the use of domain randomization in training. The code is also compatible with the OpenAI gym framework, as well as the Gazebo simulator.