Research map / Vision-language and multimodal / Visual question answering
GEMeX: A Large-Scale, Groundable, and Explainable Medical VQA Benchmark for Chest X-ray Diagnosis
Bo Liu, Ke Zou, Li-Ming Zhan, Zexin Lu, Xiaoyu Dong, Yidi Chen, Chengqiang Xie, Jiannong Cao, Xiao-Ming Wu, Huazhu Fu
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 |
|---|---|---|---|
| fine-tune LLaVA-Med-v1 on GEMeX with question-type-aware instructions | LLaVA-Med-v1 · GEMeX test (average) | average score: 72.53 -> 86.04 | Improved |
| fine-tune on GEMeX | LLaVA-Med-v1 · GEMeX closed-ended | A-score: 62.62 -> 77.35 | Improved |
| fine-tune on GEMeX | LLaVA-Med-v1 · GEMeX open-ended | AR-score: 90.34 -> 97.05 | Improved |
| fine-tune on GEMeX (enables visual grounding) | LLaVA-Med-v1 · GEMeX open-ended | V-score: - -> 51.47 | Improved |
| fine-tune on GEMeX | LLaVA-Med-v1 · GEMeX single-choice | A-score: - -> 73.08 | Improved |
| fine-tune on GEMeX versus GPT-4o-mini | LLaVA-Med-v1 · GEMeX test (average) | average score: 82.30 -> 86.04 | Improved |
| fine-tune on GEMeX versus GPT-4o-mini | LLaVA-Med-GEMeX · GEMeX | NLG metrics (average) | Improved |
| fine-tune on GEMeX and test zero-shot | LLaVA-Med-v1 · SLAKE-CXR open-ended | AR-score: 73.31 -> 82.78 | Improved |
| fine-tune on GEMeX and test zero-shot | LLaVA-Med-v1 · SLAKE-CXR closed-ended | A-score: 56.17 -> 69.79 | Improved |
| fine-tune on GEMeX and test zero-shot | LLaVA-Med-v1 · SLAKE-CXR closed-ended | AR-score: 62.35 -> 75.06 | Improved |
| use GPT-4o-mini versus multi-choice grounding | GPT-4o-mini · GEMeX multi-choice | V-score: 82.30 avg -> 19.19 V | Hurt |
| fine-tune on GEMeX | LLaVA-Med-v1 · GEMeX open-ended | BERTScore: 25.14 -> 42.69 | Improved |
Closest papers by meaning
- OmniMedVQA: A New Large-Scale Comprehensive Evaluation Benchmark for Medical LVLM (CVPR 2024 · similarity 0.90)
- BenchX: A Unified Benchmark Framework for Medical Vision-Language Pretraining on Chest X-Rays (NeurIPS 2024 · similarity 0.90)
- CXReasonBench: A Benchmark for Evaluating Structured Diagnostic Reasoning in Chest X-rays (NeurIPS 2025 · similarity 0.89)
- Falcon Medical Visual Question Answering (AAAI 2025 · similarity 0.89)
- 3D-RAD: A Comprehensive 3D Radiology Med-VQA Dataset with Multi-Temporal Analysis and Diverse Diagnostic Tasks (NeurIPS 2025 · similarity 0.89)
- Alignment, Mining and Fusion: Representation Alignment with Hard Negative Mining and Selective Knowledge Fusion for Medical Visual Question Answering (CVPR 2025 · similarity 0.89)
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.
