VinDr-CXR
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Classification | VinDR-CXR | AUC0.89 | 24 | |
| Binary disease diagnosis | VinDr-CXR OOD | Macro Accuracy93.5 | 21 | |
| Medical Image Classification | VinDr-CXR (random) | Accuracy46 | 19 | |
| Visual Attribution | VinDR-CXR MedGround-Bench-Direct | IoU14.81 | 12 | |
| Object Detection | VinDr-CXR | mAP26.5 | 12 | |
| Chest radiograph classification | VinDr-CXR (test) | AUROC0.905 | 10 | |
| Image Classification | VinDr-CXR (test) | AP (LO)19.2 | 10 | |
| Classification | VinDr-CXR (test) | AUROC95.24 | 10 | |
| Multi-label lesion detection | VinDr-CXR (test) | Precision46.7 | 9 | |
| Pixel-wise Anomaly Detection | VinDr-CXR (test) | AP24.7 | 9 | |
| Sample-wise Anomaly Detection | VinDr-CXR (test) | AP73.8 | 9 | |
| Visual Grounding | VinDr-CXR (test) | mAP@0.545.9 | 8 | |
| Chest X-ray Abnormalities Detection | VinDr-CXR | AUC78.2 | 7 | |
| Anomaly Localization | VinDr-CXR | mAP@3034.1 | 6 | |
| Pneumothorax classification | VinDR-CXR | AUC94 | 6 | |
| Phrase Grounding | VinDr-CXR | IoU46 | 6 | |
| Medical Image Classification | VinDr-CXR | AUC (Lung opacity)0.84 | 5 | |
| Learning to Defer | VinDr-CXR (test) | Balanced Accuracy85.3 | 4 | |
| Diagnostic Classification | VinDr-CXR | F1 Score54.6 | 4 | |
| Medical Image Classification | VinDr-CXR (test) | Accuracy52.37 | 3 | |
| Grounded Report Generation | VinDr-CXR (test) | F1 Score (Micro)54.6 | 3 | |
| Phrase Grounding | VinDr-CXR (test) | IoU (Min)0.244 | 2 |
Showing 22 of 22 rows