Binary Classification on COVID-BLUeS patient-level (5-fold cross-val)
69.8AccuracyMedSAM
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MedSAMClassifier=Soft binary classifier (Linear SVC), Sampling=20 frames per patient, Aggregation=Average latent vector per-patient2026.02 | 69.8 | 70.8 | 70 | 69.7 | 71 | |
| TUSAClassifier=Soft binary classifier (Linear SVC), Sampling=20 frames per patient, Aggregation=Average latent vector per-patient2026.02 | 69.8 | 70.8 | 70 | 69.7 | 75.1 | |
| USFMClassifier=Soft binary classifier (Linear SVC), Sampling=20 frames per patient, Aggregation=Average latent vector per-patient2026.02 | 62 | 64.7 | 56.7 | 66.7 | 65.4 | |
| UltraSamClassifier=Soft binary classifier (Linear SVC), Sampling=20 frames per patient, Aggregation=Average latent vector per-patient2026.02 | 60.3 | 61.5 | 60 | 60.6 | 56.2 |