AI Paper Atlas

Research map / Benchmarks / IU-Xray

IU-Xray leaderboard

26 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on IU-Xray. Best BLEU-4: 0.266 (DiA, AAAI 2026).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

IU-Xray · BLEU-4

20 papers, each with its best reported number for its own method; higher is better.

#BLEU-4 ↑MethodPaperTopic
10.266DiADiA-gnostic VLVAE: Disentangled Alignment-Constrained Vision Language Variational AutoEncoder for Robust Radiology Reporting with Missing Modalities
AAAI 2026 · Table 1
Cross-modal learning
20.225DAMPERDAMPER: A Dual-Stage Medical Report Generation Framework with Coarse-Grained MeSH Alignment and Fine-Grained Hypergraph Matching
AAAI 2025 · Table 1
Human and embodied modeling
30.208DARTDART: Disease-aware Image-Text Alignment and Self-correcting Re-alignment for Trustworthy Radiology Report Generation
CVPR 2025 · Table 1
Text-to-image diffusion
40.208OursA Self-Boosting Framework for Automated Radiographic Report Generation
CVPR 2021 · Table 1
Human and embodied modeling
50.203EKAGen (RN-101)
ResNet-101
Instance-level Expert Knowledge and Aggregate Discriminative Attention for Radiology Report Generation
CVPR 2024 · Table 1
Recognition and long-tail
60.2UAR
Transformer-Base (dVAE)
Unify, Align and Refine: Multi-Level Semantic Alignment for Radiology Report Generation
ICCV 2023 · Table 1
Cross-modal learning
70.199REVA-PO
Qwen2.5-VL-7B-Instruct
REVA-PO: Stabilizing Reinforcement Learning for Chest X-ray Report Generation
ECCV 2026 · Table 1
RL from feedback
80.199XPRONET
ResNet-101
Cross-Modal Prototype Driven Network for Radiology Report Generation
ECCV 2022 · Table 1
Cross-modal learning
90.196Ours
LLaMA3-3B + Swin
SAT-RRG: LLM-Guided Self-Adaptive Training for Radiology Report Generation with Token-Level Push-Pull Optimization
CVPR 2026 · Table 2
Language, safety and interpretability
100.187Ours (DAST+DVAF+DMSR, VMamba+Phi-4)A Disease-Aware Dual-Stage Framework for Chest X-ray Report Generation
AAAI 2026 · Table 3
Human and embodied modeling
10 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

IU-Xray · BLEU-1

14 papers, each with its best reported number for its own method; higher is better.

#BLEU-1 ↑MethodPaperTopic
10.616DiADiA-gnostic VLVAE: Disentangled Alignment-Constrained Vision Language Variational AutoEncoder for Robust Radiology Reporting with Missing Modalities
AAAI 2026 · Table 1
Cross-modal learning
20.548MPO
R2Gen
Radiology Report Generation via Multi-objective Preference Optimization
AAAI 2025 · NLG comparison table
RL from feedback
30.526EKAGen (RN-101)
ResNet-101
Instance-level Expert Knowledge and Aggregate Discriminative Attention for Radiology Report Generation
CVPR 2024 · Table 1
Recognition and long-tail
40.525KiUTKiUT: Knowledge-Injected U-Transformer for Radiology Report Generation
CVPR 2023 · Table 1
Human and embodied modeling
50.525XPRONET
ResNet-101
Cross-Modal Prototype Driven Network for Radiology Report Generation
ECCV 2022 · Table 1
Cross-modal learning
60.52DAMPERDAMPER: A Dual-Stage Medical Report Generation Framework with Coarse-Grained MeSH Alignment and Fine-Grained Hypergraph Matching
AAAI 2025 · Table 1
Human and embodied modeling
70.501Ours
DenseNet121; BERT-base
Automatic Radiology Reports Generation via Memory Alignment Network
AAAI 2024 · Table 1
Human and embodied modeling
80.499MiniGPT-4 (FT)+I3+C2FD (THIS WORK)
MiniGPT-4 (FT)
Bootstrapping Large Language Models for Radiology Report Generation
AAAI 2024 · Table 2
Language, safety and interpretability
90.495Ours
BERT-base
Clinical-BERT: Vision-Language Pre-training for Radiograph Diagnosis and Reports Generation
AAAI 2022 · Table 1
Human and embodied modeling
100.486DARTDART: Disease-aware Image-Text Alignment and Self-correcting Re-alignment for Trustworthy Radiology Report Generation
CVPR 2025 · Table 1
Text-to-image diffusion
4 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

IU-Xray · ROUGE-L

14 papers, each with its best reported number for its own method; higher is better.

#ROUGE-L ↑MethodPaperTopic
10.516DiADiA-gnostic VLVAE: Disentangled Alignment-Constrained Vision Language Variational AutoEncoder for Robust Radiology Reporting with Missing Modalities
AAAI 2026 · Table 1
Cross-modal learning
20.438CoFE
ViT-S + GPT-2 Medium
Learning with Counterfactual Explanations for Radiology Report Generation
ECCV 2024 · Table 1
Human and embodied modeling
30.416MedM2GMedM2G: Unifying Medical Multi-Modal Generation via Cross-Guided Diffusion with Visual Invariant
CVPR 2024 · Table 1
Cross-modal learning
40.415MPO
R2Gen
Radiology Report Generation via Multi-objective Preference Optimization
AAAI 2025 · NLG comparison table
RL from feedback
50.411DARTDART: Disease-aware Image-Text Alignment and Self-correcting Re-alignment for Trustworthy Radiology Report Generation
CVPR 2025 · Table 1
Text-to-image diffusion
60.411XPRONET
ResNet-101
Cross-Modal Prototype Driven Network for Radiology Report Generation
ECCV 2022 · Table 1
Cross-modal learning
70.405UAR
Transformer-Base (dVAE)
Unify, Align and Refine: Multi-Level Semantic Alignment for Radiology Report Generation
ICCV 2023 · Table 1
Cross-modal learning
80.404EKAGen (RN-101)
ResNet-101
Instance-level Expert Knowledge and Aggregate Discriminative Attention for Radiology Report Generation
CVPR 2024 · Table 1
Recognition and long-tail
90.4Ours
LLaMA3-3B + Swin
SAT-RRG: LLM-Guided Self-Adaptive Training for Radiology Report Generation with Token-Level Push-Pull Optimization
CVPR 2026 · Table 2
Language, safety and interpretability
100.397DAMPERDAMPER: A Dual-Stage Medical Report Generation Framework with Coarse-Grained MeSH Alignment and Fine-Grained Hypergraph Matching
AAAI 2025 · Table 1
Human and embodied modeling
4 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

IU-Xray · METEOR

11 papers, each with its best reported number for its own method; higher is better.

#METEOR ↑MethodPaperTopic
10.411HiMed-3B
Qwen2.5-VL-3B-Instruct
Beyond N-grams: A Hierarchical Reward Learning Framework for Clinically-Aware Medical Report Generation
AAAI 2026 · Table 3
Agents, code and math
20.316LePaX (Ours, 4B)
BLIP-3 Phi-3-mini 3.8B
Seeing What Matters: Lesion-Aware High-Resolution Patch Discovery and Fusion for Chest X-ray Report Generation
ECCV 2026 · Table 1
Multimodal LLMs
30.284DAMPERDAMPER: A Dual-Stage Medical Report Generation Framework with Coarse-Grained MeSH Alignment and Fine-Grained Hypergraph Matching
AAAI 2025 · Table 1
Human and embodied modeling
40.242KiUTKiUT: Knowledge-Injected U-Transformer for Radiology Report Generation
CVPR 2023 · Table 1
Human and embodied modeling
50.224MPO
R2Gen
Radiology Report Generation via Multi-objective Preference Optimization
AAAI 2025 · NLG comparison table
RL from feedback
60.222Ours
LLaMA3-3B + Swin
SAT-RRG: LLM-Guided Self-Adaptive Training for Radiology Report Generation with Token-Level Push-Pull Optimization
CVPR 2026 · Table 2
Language, safety and interpretability
70.22XPRONET
ResNet-101
Cross-Modal Prototype Driven Network for Radiology Report Generation
ECCV 2022 · Table 1
Cross-modal learning
80.213Ours
DenseNet121; BERT-base
Automatic Radiology Reports Generation via Memory Alignment Network
AAAI 2024 · Table 1
Human and embodied modeling
90.205DARTDART: Disease-aware Image-Text Alignment and Self-correcting Re-alignment for Trustworthy Radiology Report Generation
CVPR 2025 · Table 1
Text-to-image diffusion
100.202CoFE
ViT-S + GPT-2 Medium
Learning with Counterfactual Explanations for Radiology Report Generation
ECCV 2024 · Table 1
Human and embodied modeling
1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

IU-Xray · CIDEr

11 papers, each with its best reported number for its own method; higher is better.

#CIDEr ↑MethodPaperTopic
10.731CoFE
ViT-S + GPT-2 Medium
Learning with Counterfactual Explanations for Radiology Report Generation
ECCV 2024 · Table 1
Human and embodied modeling
20.634Ours (DAST+DVAF+DMSR, VMamba+Phi-4)A Disease-Aware Dual-Stage Framework for Chest X-ray Report Generation
AAAI 2026 · Table 3
Human and embodied modeling
30.586DCLDynamic Graph Enhanced Contrastive Learning for Chest X-Ray Report Generation
CVPR 2023 · Table 1
Knowledge and temporal graphs
40.501UAR
Transformer-Base (dVAE)
Unify, Align and Refine: Multi-Level Semantic Alignment for Radiology Report Generation
ICCV 2023 · Table 1
Cross-modal learning
50.452OursA Self-Boosting Framework for Automated Radiographic Report Generation
CVPR 2021 · Table 1
Human and embodied modeling
60.435METransformerMETransformer: Radiology Report Generation by Transformer With Multiple Learnable Expert Tokens
CVPR 2023 · Table 1
Human and embodied modeling
70.432Ours
BERT-base
Clinical-BERT: Vision-Language Pre-training for Radiograph Diagnosis and Reports Generation
AAAI 2022 · Table 1
Human and embodied modeling
80.359XPRONET
ResNet-101
Cross-Modal Prototype Driven Network for Radiology Report Generation
ECCV 2022 · Table 1
Cross-modal learning
90.351PPKEDExploring and Distilling Posterior and Prior Knowledge for Radiology Report Generation
CVPR 2021 · Table 1
Human and embodied modeling
100.343HRGR-AgentHybrid Retrieval-Generation Reinforced Agent for Medical Image Report Generation
NeurIPS 2018 · Automatic evaluation table
Human-AI and goal-conditioned
1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.