CORe50
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Out-of-Distribution Detection | Core50 ID | AUROC (COCO)98.9 | 40 | |
| Domain-incremental learning | CORe50 (test) | Test Accuracy96.7 | 34 | |
| Object Recognition | CORe50 indoor-to-outdoor sessions | Accuracy84.5 | 24 | |
| Continual Learning | CORe50 | Average Accuracy97.1 | 24 | |
| Class-Incremental Learning | CORe50 | AVG Acc90.6 | 22 | |
| Image Classification | CORe50 (test) | A_T92.29 | 22 | |
| Multi-domain generalization | CORe50 Hard | s5 Score54.4 | 18 | |
| Domain Incremental Learning | CORe50 Unknown scenarios | Average Accuracy (AA)94.37 | 15 | |
| Personalized Object Detection | CORe50 5-shot | mAP70.6 | 13 | |
| Personalized Object Detection | CORe50 1-shot | mAP60.9 | 13 | |
| Few-Shot Class-Incremental Learning | CORe50 multi-class 5-shot | BCR99.3 | 13 | |
| Few-Shot Class-Incremental Learning | CORe50 Single-class five-shot, 1 novel class | BCR99.8 | 13 | |
| Multi-class One-Shot Class-Incremental Learning | CORe50 (test) | BCR99.1 | 11 | |
| Clustering | core50 | ACC61.37 | 10 | |
| Clustering | core50 Out-of-Sample | ACC45.36 | 9 | |
| Clustering | core50 In-Sample | Accuracy61.37 | 9 | |
| Object Detection | CORe50 5 tasks | mAP@5048.7 | 8 | |
| Instance Classification | CORe50 (instance) | Accuracy71.45 | 8 | |
| Class-incremental Learning | CoRe50 10 tasks, 5 classes per task | Accuracy72.29 | 7 | |
| Image Classification | CoRe50 5 tasks, 10 classes per task | Accuracy72.96 | 7 | |
| Incremental Image Classification | CORe50 | Average Accuracy96.87 | 6 | |
| Unsupervised Continual Learning | CORe50 (test) | Latency (s)0.48 | 4 | |
| Continual Learning | CORe50 | Latency (ms)0.4769 | 3 | |
| Temporal OOD Detection | Core50 (ID) vs ImageNet-1K (OOD) (Late split t=8) | FPR957.73 | 2 | |
| Continual Object Detection | CORe50 | Metric- | 0 |
Showing 25 of 25 rows