用视觉语言模型识别家庭任务中的隐蔽安全风险，提升智能机器人判断力。
HomeGuard: VLM-based Embodied Safeguard for Identifying Contextual Risk in Household Task
- 通过分步感知与语义判断，精准定位潜在危险的交互目标和空间区域。
- 相比基础模型，风险识别准确率提升超30%，同时减少误判导致的过度保守行为。
- 适合开发具身智能机器人安全系统的研究者与工程师使用。
视觉语言模型（VLM）使具身智能体能够执行复杂指令，但在环境状态细微变化下，原本无害的指令可能引发安全隐患。现有防护机制存在不足：规则方法在物体密集场景中难以扩展，基于模型的方法依赖提示工程，易导致感知不聚焦，出现漏检或幻觉。为此，我们提出一种与架构无关的安全防护框架HomeGuard，其核心为上下文引导的思维链（CG-CoT）机制。该机制将风险评估分解为两步：首先通过逐级注意力锚定交互目标及其相关空间邻域，实现主动感知；随后基于视觉证据进行语义判断。我们构建了一个精心标注的定位数据集，并采用两阶段训练策略，利用过程奖励进行强化微调（RFT），以确保中间步骤的精确定位。实验表明，HomeGuard显著提升了安全性，在风险匹配率上优于基线模型超过30%，同时降低了过度保守现象。此外，生成的视觉锚点可作为下游规划器的显式空间约束，支持碰撞规避与安全轨迹生成。代码与数据已公开于https://github.com/AI45Lab/HomeGuard。
原文摘要 · Abstract (English)
Vision-Language Models (VLMs) empower embodied agents to execute complex instructions, yet they remain vulnerable to contextual safety risks where benign commands become hazardous due to subtle environmental states. Existing safeguards often prove inadequate. Rule-based methods lack scalability in object-dense scenes, whereas model-based approaches relying on prompt engineering suffer from unfocused perception, resulting in missed risks or hallucinations. To address this, we propose an architecture-agnostic safeguard featuring Context-Guided Chain-of-Thought (CG-CoT). This mechanism decomposes risk assessment into active perception that sequentially anchors attention to interaction targets and relevant spatial neighborhoods, followed by semantic judgment based on this visual evidence. We support this approach with a curated grounding dataset and a two-stage training strategy utilizing Reinforcement Fine-Tuning (RFT) with process rewards to enforce precise intermediate grounding. Experiments demonstrate that our model HomeGuard significantly enhances safety, improving risk match rates by over 30% compared to base models while reducing oversafety. Beyond hazard detection, the generated visual anchors serve as actionable spatial constraints for downstream planners, facilitating explicit collision avoidance and safety trajectory generation. Code and data are released under https://github.com/AI45Lab/HomeGuard
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。