Out of Distribution Detection via Domain-Informed Gaussian Process State Space Models
This work addresses safety in robot navigation for unseen environments, but it is incremental as it builds on existing GPSSM methods by incorporating domain knowledge.
The paper tackled the problem of enabling robots to safely detect out-of-distribution (OoD) scenarios online by proposing a domain-informed Gaussian process state-space model (GPSSM) and an OoD monitor, resulting in improved regression quality with smaller datasets and reliable classification of unseen terrains on a real quadruped robot.
In order for robots to safely navigate in unseen scenarios using learning-based methods, it is important to accurately detect out-of-training-distribution (OoD) situations online. Recently, Gaussian process state-space models (GPSSMs) have proven useful to discriminate unexpected observations by comparing them against probabilistic predictions. However, the capability for the model to correctly distinguish between in- and out-of-training distribution observations hinges on the accuracy of these predictions, primarily affected by the class of functions the GPSSM kernel can represent. In this paper, we propose (i) a novel approach to embed existing domain knowledge in the kernel and (ii) an OoD online runtime monitor, based on receding-horizon predictions. Domain knowledge is provided in the form of a dataset, collected either in simulation or by using a nominal model. Numerical results show that the informed kernel yields better regression quality with smaller datasets, as compared to standard kernel choices. We demonstrate the effectiveness of the OoD monitor on a real quadruped navigating an indoor setting, which reliably classifies previously unseen terrains.