mlsanityData to 14 Oct 2026

Out of Distribution Detection via Domain-Informed Gaussian Process State Space Models

arXiv 2309.06655 ↗First posted 13 Sep 2023cs.RO cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 28 of its 43 references that are in the map. First posted on arXiv 13 Sep 2023.

Builds on

28 of 43 references.

References · 28

  1. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
  2. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,348
  3. 2021Generalized Out-of-Distribution Detection: A Surveycited by 802
  4. 2018GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Accelerationcited by 609
  5. 2020Learning Invariant Representations for Reinforcement Learning without Reconstructioncited by 436
  6. 2018Gaussian Processes and Kernel Methods: A Review on Connections and Equivalencescited by 192
  7. 2017Doubly Stochastic Variational Inference for Deep Gaussian Processescited by 143
  8. 2014Hilbert Space Methods for Reduced-Rank Gaussian Process Regressioncited by 88
  9. 2016Variational Fourier features for Gaussian processescited by 85
  10. 2021Physics-Guided Deep Learning for Dynamical Systems: A Surveycited by 44
  11. 2014Variational Gaussian Process State-Space Modelscited by 43
  12. 2021Sample-Efficient Safety Assurances using Conformal Predictioncited by 41
16 more
  1. 2020Out-of-Distribution Detection for Automotive Perceptioncited by 37
  2. 2022A System-Level View on Out-of-Distribution Data in Roboticscited by 33
  3. 2021Sketching Curvature for Efficient Out-of-Distribution Detection for Deep Neural Networkscited by 32
  4. 2017Identification of Gaussian Process State Space Modelscited by 31
  5. 2019Actively Learning Gaussian Process Dynamicscited by 26
  6. 2013Bayesian Inference and Learning in Gaussian Process State-Space Models with Particle MCMCcited by 24
  7. 2019Overcoming Mean-Field Approximations in Recurrent Gaussian Process Modelscited by 21
  8. 2022Failure Prediction with Statistical Guarantees for Vision-Based Robot Controlcited by 21
  9. 2015Computationally Efficient Bayesian Learning of Gaussian Process State Space Modelscited by 18
  10. 2016A flexible state space model for learning nonlinear dynamical systemscited by 18
  11. 2020Multi-Agent Safe Planning with Gaussian Processescited by 13
  12. 2023Generative Causal Representation Learning for Out-of-Distribution Motion Forecastingcited by 13
  13. 2022The Value of Out-of-Distribution Datacited by 12
  14. 2017On the Design of LQR Kernels for Efficient Controller Learningcited by 4
  15. 2023Safety-Critical Control with Bounded Inputs via Reduced Order Modelscited by 4
  16. 2015Nonlinear State Space Model Identification Using a Regularized Basis Function Expansioncited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 1; 2026: 1.

3 papers.

  1. 2024Unpacking Failure Modes of Generative Policies: Runtime Monitoring of Consistency and Progresscited by 69
  2. 2025General OOD Detection via Model-aware and Subspace-aware Variable Prioritycited by 0
  3. 2026Kernel-Based Metrics Learning for Uncertain Opponent Vehicle Trajectory Prediction in Autonomous Racingcited by 0