AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

HomeGuard: VLM-based Embodied Safeguard for Identifying Contextual Risk in Household Task

Xiaoya Lu, Yijin Zhou, Zeren Chen, Ruocheng Wang, Bingrui Sima, Enshen Zhou, Lu Sheng, Dongrui Liu, Jing Shao

ECCV 2026 · Vision-language and multimodal › Robot action and navigation · approach: Embodied multimodal LLM agents · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add constraint IoU reward (ablation)HomeGuard-4B · HomeGuard-BenchRMR: 50.20 -> 63.72Improved
add target IoU reward (ablation)HomeGuard-4B · HomeGuard-Benchoversafety rate: 37.03 -> 21.96Improved
add target IoU reward (ablation)HomeGuard-4B · HomeGuard-BenchRMR: 56.27 -> 63.72Improved
add visual grounding (IoU) supervision overall (ablation)HomeGuard-4B · HomeGuard-Benchoversafety rate: 42.35 -> 21.96Improved
add visual grounding (IoU) supervision overall (ablation)HomeGuard-4B · HomeGuard-BenchRIR: 90.98 -> 90.00No clear effect
add RFT with CG-CoT on top of prompting only the 4-step CoT (ablation)Qwen3-VL-4B-Thinking · HomeGuard-BenchRMR: 43.50 -> 63.72Improved
train with CG-CoT instead of prompting safety principles directly (ablation)Qwen3-VL-4B-Thinking · HomeGuard-Benchoversafety rate: 35.56 -> 21.96Improved
prompt with ground-truth boxes as visual anchors (ablation)Qwen3-VL-4B-Thinking · HomeGuard-BenchRIR: 66.67 -> 84.71Improved
two-stage SFT+RFT training with CG-CoTQwen3-VL-4B-Thinking · HomeGuard-BenchRIR: 66.67 -> 90.00Improved
two-stage SFT+RFT training with CG-CoTQwen3-VL-8B-Thinking · HomeGuard-Benchoversafety rate: 32.62 -> 13.14Improved
integrate HomeGuard safety tips into plannerQwen3-VL-8B-Thinking planner · IS-Benchsafe success rate: 29.54 -> 45.65Improved
integrate HomeGuard safety tips into plannerQwen3-VL-8B-Thinking planner · IS-Benchtask success rate: 61.36 -> 73.91Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy