Image Classification on Caltech101 (Top-1 Accuracy)
95.57Top-1 Accuracy (Caltech101)SCAN
Evaluation Results
| Method | Links | |
|---|---|---|
| SCANShot=162025.05 | 95.57 | |
| GPT-4V2024.03 | 95.5 | |
| RARBase Model=Qwen-VL2024.03 | 95.5 | |
| SCANShot=82025.05 | 95 | |
| DeKgBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 94.73 | |
| MAILBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 94.73 | |
| MMRLBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 94.67 | |
| RARBase Model=LLaVA1.52024.03 | 94.6 | |
| MAIL++Backbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 94.6 | |
| CoOpOpBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 94.43 | |
| RARBase Model=Intern-IXC22024.03 | 94.4 | |
| VP (white-box)Shots=16, Model Access Type=White-box2024.07 | 94.2 | |
| LLaVA1.5 FinetuningShots=162024.03 | 94.1 | |
| RAR (LLaVA1.5)Shots=162024.03 | 94.1 | |
| TCPBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 93.97 | |
| BIP-DBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 93.93 | |
| SCANShot=42025.05 | 93.81 | |
| BARShots=16, Model Access Type=Black-box2024.07 | 93.8 | |
| MMABackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 93.8 | |
| SimNLShot=162025.05 | 93.77 | |
| BlackVIPShots=16, Model Access Type=Black-box2024.07 | 93.7 | |
| CoOpBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 93.7 | |
| PromptSRCBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 93.6 | |
| MaPLeBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 93.53 | |
| BlackVIP-SEShots=16, Model Access Type=Black-box2024.07 | 93.5 | |
| RAR (LLaVA1.5)Shots=82024.03 | 93.5 | |
| TaskResShot=162025.05 | 93.43 | |
| SimNLShot=82025.05 | 93.4 | |
| CLIP-ViT-B/16Layers=-2026.03 | 92.98 | |
| Tip-Adapter-FShot=162025.05 | 92.93 | |
| ZSEvaluation Protocol=Zero-shot2024.07 | 92.9 | |
| Col-LnLayers=3, 6, 9, KeepRate=0.72026.03 | 92.74 | |
| SCANShot=22025.05 | 92.69 | |
| CLIP-AdapterShot=162025.05 | 92.44 | |
| CLIP+KNNShots=162024.03 | 92.4 | |
| TaskResShot=82025.05 | 92.4 | |
| SimNLShot=42025.05 | 92.21 | |
| SCANShot=12025.05 | 92.17 | |
| RAR (LLaVA1.5)Shots=42024.03 | 92.1 | |
| LLaVA1.5 FinetuningShots=82024.03 | 92.1 | |
| CoOpShot=162025.05 | 91.61 | |
| Tip-Adapter-FShot=82025.05 | 91.54 | |
| CLIP-AdapterShot=82025.05 | 91.22 | |
| SimNLShot=22025.05 | 91.19 | |
| CLIP-AdapterShot=42025.05 | 90.98 | |
| TaskResShot=42025.05 | 90.97 | |
| CoCoOpShot=162025.05 | 90.9 | |
| TCA (Baseline)Layers=3, 6, 9, KeepRate=0.72026.03 | 90.87 | |
| SimNLShot=12025.05 | 90.87 | |
| Tip-Adapter-FShot=42025.05 | 90.67 | |
| CLIP+KNNShots=82024.03 | 90.6 | |
| CoCoOpShot=42025.05 | 90.31 | |
| TaskResShot=22025.05 | 90.27 | |
| CoOpShot=82025.05 | 90.15 | |
| CoCoOpShot=82025.05 | 90.14 | |
| Col-LnLayers=0, 3, 6, KeepRate=0.72026.03 | 89.9 | |
| Tip-Adapter-FShot=22025.05 | 89.81 | |
| CoCoOpShot=22025.05 | 89.47 | |
| VP w/ SPSA-GCShots=16, Model Access Type=Black-box2024.07 | 89.4 | |
| Tip-Adapter-FShot=12025.05 | 89.38 | |
| CLIP-AdapterShot=22025.05 | 89.32 | |
| RAR (LLaVA1.5)Shots=22024.03 | 89.2 | |
| CoOpShot=42025.05 | 89.17 | |
| TaskResShot=12025.05 | 88.8 | |
| CLIP-AdapterShot=12025.05 | 88.7 | |
| LLaVA1.5 FinetuningShots=42024.03 | 88.4 | |
| CoOpShot=22025.05 | 87.92 | |
| CLIP+KNNShots=42024.03 | 87.9 | |
| CoOpShot=12025.05 | 87.43 | |
| FARE2026.04 | 86.42 | |
| CLIP2026.04 | 86.19 | |
| CoCoOpShot=12025.05 | 86.01 | |
| RAR (LLaVA1.5)Shots=12024.03 | 85.2 | |
| TCA (Baseline)Layers=0, 3, 6, KeepRate=0.72026.03 | 84.67 | |
| CLIPShot=02025.05 | 84.52 | |
| PMG-AFT2026.04 | 84.5 | |
| AdvFLYPfull2026.04 | 84.4 | |
| LLaVA1.5 FinetuningShots=12024.03 | 84.1 | |
| AdvFLYP2026.04 | 83.82 | |
| CLIP+KNNShots=22024.03 | 82.9 | |
| TeCoA2026.04 | 81.31 | |
| TGA-ZSR2026.04 | 78.46 | |
| CLIP+KNNShots=12024.03 | 75.9 | |
| LLaVA1.5 FinetuningShots=22024.03 | 53.1 |