Table 8 Base ArtDL classification results
From: Multimodal prototype fusion network for paper-cut image classification
Methods | Type | Training | n-shot accuracy (%) | |
|---|---|---|---|---|
5-shot | 16-shot | |||
COOP | Prompt learning | Yes | 32.90 | 41.20 |
CoCoOp | Prompt learning | Yes | 24.20 | 38.20 |
CLIP zero-shot | Contrastive learning | No | 20.67 | 20.67 |
CLIP-linear | Deep learning | Yes | 18.90 | 25.60 |
Prototypical network | Metric learning | No | 34.95 | 36.78 |
Matching network | Metric learning | No | 30.25 | 33.85 |
AMPFN | Metric learning | Yes | 41.27 | 41.85 |
IMPFN | Metric learning | No | 39.57 | 41.37 |
