mlsanityData to 14 Oct 2026

Task-Driven Detection of Distribution Shifts with Statistical Guarantees for Robot Learning

arXiv 2106.13703 ↗First posted 25 Jun 2021cs.RO cs.LG stat.APcited by 4

Not on the map. Why?

Abstract (arXiv)

Our goal is to perform out-of-distribution (OOD) detection, i.e., to detect when a robot is operating in environments drawn from a different distribution than the ones used to train the robot. We leverage Probably Approximately Correct (PAC)-Bayes theory to train a policy with a guaranteed bound on performance on the training distribution. Our idea for OOD detection relies on the following intuition: violation of the performance bound on test environments provides evidence that the robot is operating OOD. We formalize this via statistical techniques based on p-values and concentration inequalities. The approach provides guaranteed confidence bounds on OOD detection including bounds on both the false positive and false negative rates of the detector and is task-driven and only sensitive to changes that impact the robot's performance. We demonstrate our approach in simulation and hardware for a grasping task using objects with unfamiliar shapes or poses and a drone performing vision-based obstacle avoidance in environments with wind disturbances and varied obstacle densities. Our examples demonstrate that we can perform task-driven OOD detection within just a handful of trials.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 30 of its 55 references that are in the map. First posted on arXiv 25 Jun 2021.

Builds on

30 of 55 references.

References · 30

  1. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  2. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,385
  3. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,485
  4. 2017Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networkscited by 1,328
  5. 2018Deep Anomaly Detection with Outlier Exposurecited by 942
  6. 2017Spectrally-normalized margin bounds for neural networkscited by 686
  7. 2017Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Datacited by 533
  8. 2019Fantastic Generalization Measures and Where to Find Themcited by 513
  9. 2019Scaling Out-of-Distribution Detection for Real-World Settingscited by 465
  10. 2020A Unifying Review of Deep and Shallow Anomaly Detectioncited by 425
  11. 2017Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samplescited by 389
  12. 2018Learning Confidence for Out-of-Distribution Detection in Neural Networkscited by 245
18 more
  1. 2021Testing for Outliers with Conformal p-valuescited by 134
  2. 2020Tighter risk certificates for neural networkscited by 112
  3. 2017Entropy-SGD optimizes the prior of a PAC-Bayes bound: Generalization properties of Entropy-SGD and data-dependent priorscited by 81
  4. 2019Towards neural networks that provably know when they don't knowcited by 69
  5. 2020Robust Out-of-distribution Detection for Neural Networkscited by 57
  6. 2016A Strongly Quasiconvex PAC-Bayesian Boundcited by 54
  7. 2020Real-time Out-of-distribution Detection in Learning-Enabled Cyber-Physical Systemscited by 46
  8. 2013A PAC-Bayesian Tutorial with A Dropout Boundcited by 43
  9. 2019PAC-Bayes with Backpropcited by 36
  10. 2018Open Category Detection with PAC Guaranteescited by 35
  11. 2021Sketching Curvature for Efficient Out-of-Distribution Detection for Deep Neural Networkscited by 32
  12. 2018PAC-Bayes Control: Learning Policies that Provably Generalize to Novel Environmentscited by 22
  13. 2021Provably Robust Detection of Out-of-distribution Data (almost) for freecited by 18
  14. 2020Probably Approximately Correct Vision-Based Planning using Motion Primitivescited by 16
  15. 2019Uncertainty-Based Out-of-Distribution Classification in Deep Reinforcement Learningcited by 15
  16. 2021RLAD: Time Series Anomaly Detection through Reinforcement Learning and Active Learningcited by 13
  17. 2020Policy Entropy for Out-of-Distribution Classificationcited by 11
  18. 2020Detecting Rewards Deterioration in Episodic Reinforcement Learningcited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 2.

4 papers.

  1. 2026Confidence-Gated Robot Autonomy: When Does Uncertainty Actually Help?cited by 1
  2. 2025COOkeD: Ensemble-based OOD detection in the era of zero-shot CLIPcited by 0
  3. 2025Deceptive Risk Minimization: Out-of-Distribution Generalization by Deceiving Distribution Shift Detectorscited by 0
  4. 2026Autonomous Drift Learning in Data Streams: A Unified Perspectivecited by 0