ProgressVLA: 用于视觉-语言机器人操控的进展引导扩散策略
ProgressVLA: Progress-Guided Diffusion Policy for Vision-Language Robotic Manipulation
AI总结：
本文提出ProgressVLA模型，通过鲁棒的进展估计和可微进展引导方法提升机器人操控任务的成功率和泛化能力。
摘要：
现有视觉-语言-动作(VLA)模型在机器人操控中缺乏进展意识，通常依赖手工设计的启发式方法来判断任务终止。本文研究了任务进展的估计与整合，提出了一种名为ProgressVLA的新模型。我们的技术贡献包括：(1) 鲁棒的进展估计：我们在大规模无监督视频-文本机器人数据集上预训练了一个进展估计器，该估计器在模拟中实现了低预测残差(在[0,1]范围内为0.07)，并在未见过的现实样本上展示了零样本泛化能力；(2) 可微进展引导：我们引入了一个逆动力学世界模型，将预测的动作令牌映射到未来的潜在视觉状态。这些潜在状态随后由进展估计器处理；通过应用最大进展正则化，我们建立了可微的管道，为动作令牌提供进展引导以优化动作。在CALVIN和LIBERO基准测试以及现实机器人部署中，广泛实验一致显示了在成功率和泛化能力上显著优于强基线模型的改进。
英文摘要：
Most existing vision-language-action (VLA) models for robotic manipulation lack progress awareness, typically relying on hand-crafted heuristics for task termination. This limitation is particularly severe in long-horizon tasks involving cascaded sub-goals. In this work, we investigate the estimation and integration of task progress, proposing a novel model named {\textbf \vla}. Our technical contributions are twofold: (1) \emph{robust progress estimation}: We pre-train a progress estimator on large-scale, unsupervised video-text robotic datasets. This estimator achieves a low prediction residual (0.07 on a scale of $[0, 1]$) in simulation and demonstrates zero-shot generalization to unseen real-world samples, and (2) \emph{differentiable progress guidance}: We introduce an inverse dynamics world model that maps predicted action tokens into future latent visual states. These latents are then processed by the progress estimator; by applying a maximal progress regularization, we establish a differentiable pipeline that provides progress-piloted guidance to refine action tokens. Extensive experiments on the CALVIN and LIBERO benchmarks, alongside real-world robot deployment, consistently demonstrate substantial improvements in success rates and generalization over strong baselines.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
