# RoboFQE：从离线轨迹无 Rollout 评估视觉机器人策略

> 原文链接：https://robotics.hellobike.com/papers/2026-10-13-2610-18699
> 频道：papers | 更新日期：2026-10-13 | 标签：arxiv, cs.RO
> 核心摘要：RoboFQE用离线机器人轨迹进行无Rollout离策略评估，在仿真与真实操纵任务中验证策略回报估计。

# RoboFQE：从离线轨迹无 Rollout 评估视觉机器人策略

::: v-pre
> 原文链接：[RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories](https://arxiv.org/abs/2610.18699)（arXiv 2610.18699 · 发布日期：2026-10-13）

**RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories**

RoboFQE用离线机器人轨迹进行无Rollout离策略评估，在仿真与真实操纵任务中验证策略回报估计。

## 核心要点

- 部署前评估机器人策略通常依赖真实执行、仿真器或世界模型产生新轨迹。
- RoboFQE结合拟合Q评估、预训练视觉基础模型和学习式终端成功标注估计折扣期望回报。
- 在仿真和真实操纵任务中，RoboFQE能准确估计折扣回报并保持策略排序。
- 消融显示预训练视觉编码器优于从头训练，学习式终端标注仅轻微降低准确性。

**验证范围与局限**：验证范围为仿真与真实世界操纵任务，包括长时程操纵任务。

## 英文摘要 (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 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 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 substantially 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.

作者：Yunfu Deng, Sangwoo Shin, Brahma S. Pavse, Josiah P. Hanna
:::

## 原始参考来源 (Sources)

- [arXiv 2610.18699：RoboFQE: Rollout-Free Evaluation of Visual Robot Policies from Offline Trajectories](https://arxiv.org/abs/2610.18699)