Image Classification on Car (Accuracy, ECE)
81.7AccuracyDataDream
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DataDreamEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.02 | 81.7 | — | — | |
| ChimeraLoRAEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.02 | 79.6 | — | — | |
| LoFTEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.02 | 78 | — | — | |
| CLIPEvaluation Protocol=Few-shot (Real-only), Shot Count=42026.02 | 74.3 | — | — | |
| IsSynthEvaluation Protocol=Few-shot (Augmented), Shot Count=4 Real + 500 Synthetic2026.02 | 71.5 | — | — | |
| CLIPEvaluation Protocol=Zero-shot, Shot Count=02026.02 | 65.2 | — | — | |
| TPTBackbone=CLIP-RN502025.03 | 58 | 3.76 | — | |
| C-TPTBackbone=CLIP-RN502025.03 | 56.5 | 1.94 | — | |
| O-TPTBackbone=CLIP-RN502025.03 | 56.44 | 1.69 | — | |
| Robust-adapt-Penalty-CTPTBackbone=CLIP-RN502025.03 | 56.4 | 1.99 | — | |
| Robust-adapt-SaLs-CTPTBackbone=CLIP-RN502025.03 | 56.27 | 2.53 | — | |
| Robust-adapt-ZS-CTPTBackbone=CLIP-RN502025.03 | 56.26 | 4.85 | — | |
| Zero ShotBackbone=CLIP-RN502025.03 | 55.7 | 4.7 | — | |
| C-TPTBackbone=CLIP-B/162025.03 | — | — | 22 | |
| O-TPTBackbone=CLIP-B/162025.03 | — | — | 107 | |
| TPTBackbone=CLIP-B/162025.03 | — | — | 25 | |
| Zero ShotBackbone=CLIP-B/162025.03 | — | — | 23 |