Generalized Category Discovery on SoyAgeing-R3 (train)
50.87Overall AccuracyRAPL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RAPLBackbone=ResNet-50, Pre-training=ImageNet2024.05 | 50.87 | 58.47 | 47.07 | |
| RAPLEvaluation Protocol=task-aware, Backbone=ResNet-502024.05 | 46.06 | — | — | |
| GCDBackbone=ResNet-50, Pre-training=ImageNet2024.05 | 38.64 | 47.73 | 34.09 | |
| IICEvaluation Protocol=task-aware, Backbone=ResNet-502024.05 | 34.55 | — | — | |
| UNOEvaluation Protocol=task-aware, Backbone=ResNet-502024.05 | 33.37 | — | — | |
| XConBackbone=ResNet-50, Pre-training=ImageNet2024.05 | 33.08 | 42.68 | 28.28 | |
| rKDEvaluation Protocol=task-aware, Backbone=ResNet-502024.05 | 32.73 | — | — | |
| ComExEvaluation Protocol=task-aware, Backbone=ResNet-502024.05 | 32.32 | — | — | |
| RankStats ILEvaluation Protocol=task-aware, Backbone=ResNet-502024.05 | 14.55 | — | — |