Counterfactual Behavior Cloning: Offline Imitation Learning from Imperfect Human Demonstrations
ACM Transactions on Human-Robot Interaction · 30.11.2026 · 15 (6) · 1-24
DOI: 10.1145/3842741 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Human-Robot Interaction, online erschienen am 30.10.2026. Verfasst von Shahabedin Sagheb und 1 weiteren Autor:innen. Die Arbeit nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Science Foundation.
Abstract (Originalsprache)
Learning from humans is challenging because people are imperfect teachers. When everyday humans show the robot a new task they want it to perform, humans inevitably make errors (e.g., inputting noisy actions) and provide suboptimal examples (e.g., overshooting the goal). Existing methods often learn by matching some or all of the human’s behavior—but this approach is fundamentally limited because the demonstrations themselves are imperfect. In this work, we advance offline imitation learning by enabling robots to extrapolate across nearby actions, instead of only considering what the human actually showed. We achieve this by hypothesizing that all of the human’s demonstrations are trying to convey an underlying policy, while the noise and sub-optimality within their behaviors obfuscate the data and introduce unintentional complexity. To recover the underlying policy and learn what the human teacher meant, we introduce Counter-BC , a generalized version of behavior cloning. Counter-BC expands the given dataset to include actions close to behaviors the human demonstrated (i.e., counterfactual actions that the human teacher could have intended, but did not actually show). During training, Counter-BC autonomously modifies the human’s demonstrations within this expanded region to reach a simplified policy that explains the underlying trends in the human’s dataset. Theoretically, we prove that Counter-BC can extract a simple and similar-to-demonstration policy from imperfect data, multiple users, and teachers of varying skill levels. Empirically, we compare Counter-BC to state-of-the-art alternatives in simulated and real-world settings with noisy demonstrations, standardized datasets, and real human teachers. Overall, we find that trying to extrapolate what the human teacher meant by considering nearby actions—as opposed to rigorously following what actions the human actually demonstrated—can lead to more proficient learning from humans. See videos of our work here: https://youtu.be/XaeOZWhTt68 .
Förderung
- National Science Foundation
Zitieren
APA
Sagheb, S., & Losey, D. P. (2026). Counterfactual Behavior Cloning: Offline Imitation Learning from Imperfect Human Demonstrations. ACM Transactions on Human-Robot Interaction, 15(6), 1-24. https://doi.org/10.1145/3842741
Vancouver
Sagheb S, Losey DP. Counterfactual Behavior Cloning: Offline Imitation Learning from Imperfect Human Demonstrations. ACM Transactions on Human-Robot Interaction. 2026;15(6):1-24. doi:10.1145/3842741
BibTeX
@article{Sagheb2026,
author = {Sagheb, Shahabedin and Losey, Dylan P.},
title = {Counterfactual Behavior Cloning: Offline Imitation Learning from Imperfect Human Demonstrations},
journal = {ACM Transactions on Human-Robot Interaction},
year = {2026},
volume = {15},
number = {6},
pages = {1--24},
doi = {10.1145/3842741}
}
EUVIMED