Object Detection on Object-CXR
22.9mAPSISTA
Evaluation results
| Method | Links | |
|---|---|---|
| SISTAFine-tuning percentage=10%2026.01 | 22.9 | |
| Med-UniCTraining Data Fraction=100%, Disease-level annotations=false2023.05 | 21.6 | |
| SISTAFine-tuning percentage=100%2026.01 | 20.9 | |
| G2D (encoder)Training data fraction=100%, architecture=encoder2023.12 | 20.4 | |
| MLIPFine-tuning percentage=100%, Auxiliary knowledge=true2026.01 | 20.2 | |
| PRIORFine-tuning percentage=100%2026.01 | 19.8 | |
| M-FLAGFine-tuning percentage=100%2026.01 | 19.5 | |
| M-FLAGTraining data fraction=100%2023.12 | 19.3 | |
| MGCATraining Data Fraction=100%, Disease-level annotations=true2023.05 | 19.2 | |
| MGCATraining data fraction=100%2023.12 | 19.2 | |
| MGCAFine-tuning percentage=100%2026.01 | 19.2 | |
| MLIPFine-tuning percentage=10%, Auxiliary knowledge=true2026.01 | 17.4 | |
| GLoRIA-MIMICTraining Data Fraction=100%, Disease-level annotations=false2023.05 | 16.6 | |
| GLoRIA-MIMICTraining data fraction=100%2023.12 | 16.6 | |
| GLoRIAFine-tuning percentage=100%2026.01 | 16.6 | |
| ConVIRTTraining Data Fraction=100%, Disease-level annotations=false2023.05 | 15.9 | |
| ConVIRTTraining data fraction=100%2023.12 | 15.9 | |
| ImageNetFine-tuning percentage=100%2026.01 | 15.9 | |
| ConVIRTFine-tuning percentage=100%2026.01 | 15.9 | |
| GLoRATraining Data Fraction=100%, Disease-level annotations=false2023.05 | 15.6 | |
| GLoRATraining data fraction=100%2023.12 | 15.6 | |
| PRIORFine-tuning percentage=10%2026.01 | 15.2 | |
| M-FLAGFine-tuning percentage=10%2026.01 | 13.6 | |
| Med-UniCTraining Data Fraction=10%, Disease-level annotations=false2023.05 | 13.3 | |
| G2D (encoder)Training data fraction=10%, architecture=encoder2023.12 | 13.1 | |
| M-FLAGTraining data fraction=10%2023.12 | 12.4 | |
| MGCATraining Data Fraction=10%, Disease-level annotations=true2023.05 | 12.1 | |
| MGCATraining data fraction=10%2023.12 | 12.1 | |
| MGCAFine-tuning percentage=10%2026.01 | 12.1 | |
| MedKLIPTraining Data Fraction=100%, Disease-level annotations=true2023.05 | 11.6 | |
| MedKLIP*Training data fraction=100%, annotations=disease-level2023.12 | 11.6 | |
| MedKLIPFine-tuning percentage=100%, Instance-level labels=true2026.01 | 11.6 | |
| SISTAFine-tuning percentage=1%2026.01 | 11 | |
| GLoRATraining Data Fraction=10%, Disease-level annotations=false2023.05 | 10.6 | |
| GLoRATraining data fraction=10%2023.12 | 10.6 | |
| GLoRIA-MIMICTraining Data Fraction=10%, Disease-level annotations=false2023.05 | 8.9 | |
| GLoRIA-MIMICTraining data fraction=10%2023.12 | 8.9 | |
| GLoRIAFine-tuning percentage=10%2026.01 | 8.9 | |
| ConVIRTTraining Data Fraction=10%, Disease-level annotations=false2023.05 | 8.6 | |
| ConVIRTTraining data fraction=10%2023.12 | 8.6 | |
| ImageNetFine-tuning percentage=10%2026.01 | 8.6 | |
| ConVIRTFine-tuning percentage=10%2026.01 | 8.6 | |
| ImageNetTraining Data Fraction=100%, Disease-level annotations=false2023.05 | 8.3 | |
| ImageNet InitTraining data fraction=100%2023.12 | 8.3 | |
| MedKLIPTraining Data Fraction=10%, Disease-level annotations=true2023.05 | 7.1 | |
| MedKLIP*Training data fraction=10%, annotations=disease-level2023.12 | 7.1 | |
| MedKLIPFine-tuning percentage=10%, Instance-level labels=true2026.01 | 7.1 | |
| Med-UniCTraining Data Fraction=1%, Disease-level annotations=false2023.05 | 6.6 | |
| MLIPFine-tuning percentage=1%, Auxiliary knowledge=true2026.01 | 4.6 | |
| RandomTraining Data Fraction=100%, Disease-level annotations=false2023.05 | 4.4 | |
| Random InitTraining data fraction=100%2023.12 | 4.4 | |
| RandomFine-tuning percentage=100%2026.01 | 4.4 | |
| G2D (encoder)Training data fraction=1%, architecture=encoder2023.12 | 3.8 | |
| ImageNetTraining Data Fraction=10%, Disease-level annotations=false2023.05 | 2.9 | |
| ImageNet InitTraining data fraction=10%2023.12 | 2.9 | |
| PRIORFine-tuning percentage=1%2026.01 | 2.9 | |
| RandomTraining Data Fraction=10%, Disease-level annotations=false2023.05 | 0.5 | |
| Random InitTraining data fraction=10%2023.12 | 0.5 |