Fine-grained Image Classification on StanfordCars base classes
78.65AccuracyTemp. Scal.
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Temp. Scal.Prompt Strategy=CoOp2026.02 | 78.65 | 6.63 | |
| ZS-NormPrompt Strategy=CoOp2026.02 | 77.78 | 11.3 | |
| TCPTPrompt Strategy=CoOp2026.02 | 77.32 | 7.1 | |
| PenaltyPrompt Strategy=CoOp2026.02 | 77.05 | 10.01 | |
| CoOp2026.02 | 76.22 | 3.73 | |
| MBLSPrompt Strategy=CoOp2026.02 | 76.21 | 12.2 | |
| MBLSPrompt Strategy=KGCoOp2026.02 | 75.34 | 13.43 | |
| ZS-NormPrompt Strategy=KGCoOp2026.02 | 74.55 | 3.85 | |
| TCPTBackbone=ViT-B/322026.02 | 73.78 | 6.77 | |
| PenaltyBackbone=ViT-B/322026.02 | 73.33 | 10.54 | |
| CoOpBackbone=ViT-B/322026.02 | 73.13 | 4.86 | |
| ZS-NormPrompt Strategy=MaPLe2026.02 | 73.07 | 8.66 | |
| MaPLe2026.02 | 72.93 | 7.25 | |
| ZS-NormBackbone=ViT-B/322026.02 | 72.86 | 9.7 | |
| TCPTPrompt Strategy=MaPLe2026.02 | 72.8 | 7.92 | |
| MBLSPrompt Strategy=MaPLe2026.02 | 72.77 | 19.06 | |
| Temp. Scal.Prompt Strategy=MaPLe2026.02 | 72.7 | 4.96 | |
| KGCoOp2026.02 | 72.7 | 10.16 | |
| PenaltyPrompt Strategy=KGCoOp2026.02 | 72.45 | 10.58 | |
| PenaltyPrompt Strategy=MaPLe2026.02 | 72.43 | 13.53 | |
| TCPTPrompt Strategy=KGCoOp2026.02 | 71.65 | 8.1 | |
| CoOpBackbone=RN-502026.02 | 70.22 | 6.15 | |
| Temp. Scal.Prompt Strategy=KGCoOp2026.02 | 70.08 | 11.7 | |
| TCPTBackbone=RN-502026.02 | 69.9 | 8.65 | |
| ZS-NormBackbone=RN-502026.02 | 69.89 | 10.52 | |
| PenaltyBackbone=RN-502026.02 | 68.58 | 11.85 | |
| Zero Shot2026.02 | 63.6 | 3.74 |