mlsanityData to 7 Oct 2026

Discriminative Barrier Functions for Safe Adversarial Imitation Learning from Observation

arXiv 2607.13938 ↗First posted 15 Jul 2026cs.ROcited by 0

On the map: Offline RL and imitation learning from data → Safe offline RL

Abstract (arXiv)

Inverse Reinforcement Learning (IRL) algorithms are powerful tools for learning from and generalizing expert demonstrations, but they often rely on unconstrained exploration, rendering them unsafe for real-world deployment. Meanwhile, Control Barrier Functions (CBFs) can guarantee the safety of control systems, but the analytical design of CBFs can be time-consuming and esoteric. In this work, we address these limitations jointly by constraining reward function candidacy during IRL to the space of CBFs, yielding a formulation that exhibits safe online control with continuous experiential improvement. Crucially, this framework enables the data-driven recovery of barrier functions directly from unlabeled expert observations. We demonstrate that the recovered barrier function is robust to unsafe states entirely absent from the expert data. Furthermore, we benchmark our method against standard IRL baselines in a simulated navigation environment, demonstrating improved safety performance. Finally, we investigate the trade-offs of planning-based versus policy-based IRL methods across both simulation and a real world obstacle avoidance task.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Safe offline RL, one of 13 directions in Offline RL and imitation learning from data. Builds on 22 of its 43 references that are in the map. First posted on arXiv 15 Jul 2026.

Builds on

22 of 43 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025V-OCBF: Learning Safety Filters from Offline Data via Value-Guided Offline Control Barrier Functions11 Dec
  2. 2022Safe Inverse Reinforcement Learning via Control Barrier Function6 Dec
  3. 2020Learning Control Barrier Functions from Expert Demonstrations7 Apr

Other references · 19

  1. 2017Improved Training of Wasserstein GANscited by 2,223
  2. 2016Generative Adversarial Imitation Learningcited by 1,576
  3. 2018Behavioral Cloning from Observationcited by 482
  4. 2019Control Barrier Functions: Theory and Applicationscited by 332
  5. 2021What Matters for Adversarial Imitation Learning?cited by 70
  6. 2023How to Train Your Neural Control Barrier Function: Learning Safety Filters for Complex Input-Constrained Systemscited by 40
  7. 2020Synthesis of Control Barrier Functions Using a Supervised Machine Learning Approachcited by 35
  8. 2024GCBF+: A Neural Graph Control Barrier Function Framework for Distributed Safe Multi-Agent Controlcited by 20
  9. 2019MuSHR: A Low-Cost, Open-Source Robotic Racecar for Education and Researchcited by 17
  10. 2023In-Distribution Barrier Functions: Self-Supervised Policy Filters that Avoid Out-of-Distribution Statescited by 15
  11. 2023Learning Risk-Aware Costmaps via Inverse Reinforcement Learning for Off-Road Navigationcited by 15
  12. 2024Neural Control Barrier Functions for Safe Navigationcited by 8
7 more
  1. 2025CBF-RL: Safety Filtering Reinforcement Learning in Training with Control Barrier Functionscited by 8
  2. 2025How to Train Your Latent Control Barrier Function: Smooth Safety Filtering Under Hard-to-Model Constraintscited by 7
  3. 2022Walking in Narrow Spaces: Safety-critical Locomotion Control for Quadrupedal Robots with Duality-based Optimizationcited by 6
  4. 2024Learning Control Barrier Functions and their application in Reinforcement Learning: A Surveycited by 3
  5. 2023Risk-Sensitive Inhibitory Control for Safe Reinforcement Learningcited by 2
  6. 2025Model Predictive Adversarial Imitation Learning for Planning from Observationcited by 2
  7. 2023Lipschitz Continuity of Signal Temporal Logic Robustness Measures: Synthesizing Control Barrier Functions from One Expert Demonstrationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Individual Control Barrier Functions-Guided Diffusion Model for Safe Offline Multi-Agent Reinforcement Learning
  2. 2022End-to-End Imitation Learning with Safety Guarantees using Control Barrier Functions
  3. 2020oIRL: Robust Adversarial Inverse Reinforcement Learning with Temporally Extended Actions
  4. 2020f-IRL: Inverse Reinforcement Learning via State Marginal Matching
  5. 2020Bayesian Robust Optimization for Imitation Learning
  6. 2025Learning Conservative Neural Control Barrier Functions from Offline Data
  7. 2019Wasserstein Adversarial Imitation Learning
  8. 2026TAIL-Safe: Task-Agnostic Safety Monitoring for Imitation Learning Policies