Fine-grained species classification on Mollusca Species196 16-shot (test)
71.6AccuracyCLAP + POC
Evaluation Results
| Method | Links | |
|---|---|---|
| CLAP + POCstrategy=linear probing, variant=POC, shots=162025.12 | 71.6 | |
| Few-shot FT + POCstrategy=finetuning, variant=POC, shots=162025.12 | 69.2 | |
| CrossModal LP + POCstrategy=linear probing, variant=POC, shots=162025.12 | 65.2 | |
| Few-shot LP + POCstrategy=linear probing, variant=POC, shots=162025.12 | 63.9 | |
| Few-shot FTstrategy=finetuning, variant=expert, shots=162025.12 | 63.6 | |
| CLAPstrategy=linear probing, variant=expert, shots=162025.12 | 63.5 | |
| CoOp + POCstrategy=prompt learning, variant=POC, shots=162025.12 | 59.7 | |
| FineR + POCstrategy=non-learned, variant=POC, shots=162025.12 | 58.7 | |
| CrossModal LPstrategy=linear probing, variant=expert, shots=162025.12 | 58.2 | |
| CLIP-Adapter + POCstrategy=adapter learning, variant=POC, shots=162025.12 | 57.4 | |
| Few-shot LPstrategy=linear probing, variant=expert, shots=162025.12 | 56.8 | |
| CLIP-Adapterstrategy=adapter learning, variant=expert, shots=162025.12 | 48.4 | |
| FineRstrategy=non-learned, variant=expert, shots=162025.12 | 47.7 | |
| Tip-Adapter + POCstrategy=adapter learning, variant=POC, shots=162025.12 | 44.7 | |
| CoOpstrategy=prompt learning, variant=expert, shots=162025.12 | 44.5 | |
| MaPLe + POCstrategy=prompt learning, variant=POC, shots=162025.12 | 42.9 | |
| Tip-Adapterstrategy=adapter learning, variant=expert, shots=162025.12 | 27.3 | |
| MaPLestrategy=prompt learning, variant=expert, shots=162025.12 | 23.6 |