论文/arXiv 2609.33464·1 个月前

VIDEAS：通过先验引导仿真从演示视频中为世界模型蒸馏显式动作语义

VIDEAS将演示视频蒸馏为显式动作语义，经先验引导仿真验证，用于训练并评测语言世界模型VIDEAS-WM。

AI 整理

• 本体 SDK 与控制
• 仿真与训练
• 数据与采集

A智元精灵 G2

原文链接：VIDEAS: Distilling Explicit Action Semantics from Demonstration Videos for World Models via Prior-Guided Simulation（arXiv 2609.33464 · 发布日期：2026-09-27）

VIDEAS: Distilling Explicit Action Semantics from Demonstration Videos for World Models via Prior-Guided Simulation

VIDEAS将演示视频蒸馏为显式动作语义，经先验引导仿真验证，用于训练并评测语言世界模型VIDEAS-WM。

核心要点

• 动机是让世界模型显式推理动作前提与后续状态转移，而不只预测未来状态。
• 方法将视觉演示分解为离散动作轨迹，并用VLM提取动作前提和效果的结构化知识。
• 先验引导的文本环境轨迹仿真用于验证物理一致性，并加入负轨迹丰富知识与数据多样性。
• VIDEAS-WM以AgiBot-World衍生的34K高质量样本训练，在高层具身动作语义推理中达到SOTA表现。

验证范围与局限：验证范围为基于AgiBot-World衍生样本训练的高层具身动作语义推理实验及未见场景泛化。

英文摘要 (Abstract)

World models learn internal representations of environment dynamics to predict future states, enabling agents to optimize action plans without physical interactions. However, developing world models that genuinely internalize underlying causal physical laws to explicitly reason about action preconditions and subsequent state transitions remains an open challenge. In this paper, we propose VIDEAS, a data distillation framework that transforms continuous physical dynamics from operational videos into explicit action semantics for foundation models. Specifically, it deconstructs visual demonstrations into discrete action trajectories and utilizes advanced vision-language models (VLMs) to extract structured knowledge encapsulating action preconditions and effects. To ensure physical consistency, we introduce a prior-guided trajectory simulation mechanism grounded within a text-based environment to rigorously validate the extracted knowledge. Notably, we incorporate negative trajectories to enrich knowledge completeness and enhance data diversity to mitigate cognitive bias. Furthermore, we present VIDEAS-WM, an 8B/9B-parameter suite of language-based world models trained on 34K high-quality samples derived from AgiBot-World dataset. Extensive experiments demonstrate that VIDEAS-WM establishes state-of-the-art performance in high-level embodied action semantic reasoning, exhibiting profound physical understanding and robust generalization across unseen scenarios.

作者：Jianan Wang, Haoquan Zhai, Siyang Zhang, Bin Li, Juan Chen, Jingtao Qi, Zhuo Zhang, Enze Wang, Haoxiang Jin, Chen Qian

来源

本页事实对应的公开页面，以原文为准。

1. arXiv 2609.33464：VIDEAS: Distilling Explicit Action Semantics from Demonstration Videos for World Models via Prior-Guided Simulation访问 2026-10-21