RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories
RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories
This research presents RoboFQE, an off-policy evaluation framework that estimates the discounted expected return of target robot policies using offline trajectories without requiring new rollouts or behavior-policy density estimation.
The paper introduces RoboFQE, a method for off-policy evaluation of visual robot policies using offline trajectories without requiring new rollouts or behavior-policy density estimation; experiments in simulated and real-world settings demonstrate accurate return estimates and preserved policy rankings on long-horizon tasks where video world-model evaluators degrade due to prediction errors.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Data origin
- Robot trajectoriesOpen source S3
Dataset details reported by the source
- Modalities
- vision foundation model Open source S3
- Tasks
- visual manipulation Open source S4
- Embodiments
- simulated and real-world Open source S4
- paper: not_reported
What remains unknown
Dataset licensing and commercial-use rights are not established here.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
off-policy evaluation as an alternative to rollout-based robot-policy evaluation methods. Specifically, we introduce RoboFQE, an evaluator that estimates the discounted expected return of a target policy from offline robot trajectories without estimating a behavior-policy density, combining fitted Q-evaluation with a pretrained vision foundation model and learned terminal success annotation to bridge the gap between
Open source S3
standard OPE benchmarks and continuous-control visual manipulation. Controlled experiments across simulated and real-world manipulation tasks show that RoboFQE produces accurate discounted-return estimates and preserves policy rankings even on long-horizon manipulation tasks where video world-model evaluators are degraded by autoregressive prediction errors. Ablation experiments show that pretrained vision encoders s
Open source S4
arXiv:2610.18699v2 Announce Type: replace Abstract: Evaluating robot policies before deployment typically requires producing new trajectories through real-world execution, simulators, or learned world models. These rollout-based approaches impose infrastructure requirements; simulator and world-model evaluators additionally face prediction errors that can compound over the evaluation horizon. In this paper, we study
Open source S2
ubstantially outperform from-scratch alternatives, particularly on long-horizon tasks, and that learned terminal success annotation enables evaluation without ground-truth reward labels with only slight reduction of accuracy.
Open source S5
RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories
Open source S1
Implications for data suppliers
RoboSignal interpretation and collection questions, not statements of buyer demand.
- Match the buyer’s robot embodiment, action space, sensors and timestamp requirements before planning collection.
- Check the exact dataset version, license, modalities and task coverage against the intended use.
- Validate demand and acceptance criteria with a buyer before scaling. Publication, popularity and a research result do not establish a purchase commitment.
Source：arXiv Robotics — research abstracts · arxiv.org