Few-shot classification on ImageNet
78.1Base AccuracyAAPE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| AAPEPrompting Strategy=LLM-based2024.10 | 78.1 | 71.98 | 74.92 | |
| PromptSRCPrompting Strategy=Prompt learning without language priors2024.10 | 77.6 | 70.73 | 74.01 | |
| CLIPoodPrompting Strategy=Prompt learning without language priors2024.10 | 77.5 | 70.3 | 73.7 | |
| OGENPrompting Strategy=Prompt learning without language priors2024.10 | 77.5 | 70.97 | 74.09 | |
| ProGradPrompting Strategy=Human-generated prompts2024.10 | 77.02 | 66.66 | 71.46 | |
| MaPLePrompting Strategy=Prompt learning without language priors2024.10 | 76.66 | 70.54 | 73.47 | |
| CoOpPrompting Strategy=Prompt learning without language priors2024.10 | 76.47 | 67.88 | 71.92 | |
| LASP-VPrompting Strategy=Human-generated prompts2024.10 | 76.25 | 71.17 | 73.62 | |
| CoCoOpPrompting Strategy=Prompt learning without language priors2024.10 | 75.98 | 70.43 | 73.1 | |
| KgCoOpPrompting Strategy=Human-generated prompts2024.10 | 75.83 | 69.96 | 72.78 | |
| CuPLPrompting Strategy=LLM-based2024.10 | 75.05 | 68.43 | 71.59 |