Research map / Vision-language and multimodal / Visual question answering
BridgeEQA: Virtual Embodied Agents for Real Bridge Inspections
Subin Varghese, Joshua Gao, Asad Ur Rahman, Vedhus Hoskere
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace Multi-Frame VLM with EMVR (Images + SG) navigation | Grok 4 Fast · BridgeEQA | Answer Correctness: 0.576 -> 0.648 | Improved |
| replace Multi-Frame VLM with EMVR (Images + SG) | Grok 4 Fast · BridgeEQA | Image Citation Relevance: 0.687 -> 0.889 | Improved |
| replace Multi-Frame VLM with EMVR (Images + SG) | Gemini 2.5 Flash Lite · BridgeEQA | Answer Correctness: 0.507 -> 0.497 | Hurt |
| replace Multi-Frame VLM with EMVR (Images + SG) | Gemini 2.5 Flash Lite · BridgeEQA | Image Citation Relevance: 0.717 -> 0.849 | Improved |
| replace Multi-Frame VLM with EMVR (SG Only) | Gemini 2.5 Flash · BridgeEQA | Answer Correctness: 0.484 -> 0.609 | Improved |
| add scene graph to Multi-Frame VLM context (ablation) | Multi-Frame VLM · BridgeEQA | Answer Correctness: 0.507 -> 0.581 (Flash Lite); 0.576 -> 0.622 (Grok) | Improved |
| add scene graph to Multi-Frame VLM context (ablation) | Multi-Frame VLM · BridgeEQA | Image Citation Relevance: 0.687 -> 0.833 (Grok) | Improved |
| add images to EMVR initial context (SG Only -> Images + SG) (ablation) | EMVR · BridgeEQA | Answer Correctness: 0.609 -> 0.551 (Gemini Flash); 0.638 -> 0.648 (Grok) | Mixed |
| add images to EMVR initial context (ablation) | EMVR · BridgeEQA | Image Citation Relevance: 0.836 -> 0.803 (Gemini Flash); 0.876 -> 0.889 (Grok) | Mixed |
| EMVR vs Socratic LLM w/ SG | EMVR (Images + SG) · BridgeEQA | Answer Correctness: 0.542 -> 0.497 (Flash Lite); 0.614 -> 0.648 (Grok) | Mixed |
Closest papers by meaning
- OpenEQA: Embodied Question Answering in the Era of Foundation Models (CVPR 2024 · similarity 0.90)
- Extending Embodied Question Answering from Perception to Decision (CVPR 2026 · similarity 0.89)
- IndustryEQA: Pushing the Frontiers of Embodied Question Answering in Industrial Scenarios (NeurIPS 2025 · similarity 0.89)
- Embodied Scene Understanding for Vision Language Models via MetaVQA (CVPR 2025 · similarity 0.88)
- Multi-modal Situated Reasoning in 3D Scenes (NeurIPS 2024 · similarity 0.87)
- ECBench: Can Multi-modal Foundation Models Understand the Egocentric World? A Holistic Embodied Cognition Benchmark (CVPR 2025 · similarity 0.86)
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.
