Image Classification on OxfordPets
95.9AccuracyPiSSA
Evaluation Results
| Method | Links | |
|---|---|---|
| PiSSABackbone=ViT-Base, Rank=322026.05 | 95.9 | |
| PiSSABackbone=ViT-Large, Memory=1229.9 MB2026.05 | 95.7 | |
| LAGOBackbone=L/14, Zero-shot=true2026.05 | 95.07 | |
| BiFTABackbone=L/14, Zero-shot=true2026.05 | 94.96 | |
| CERSABackbone=ViT-Large, Memory=1232.3 MB2026.05 | 94.9 | |
| FourierFTBackbone=ViT-Large, Method=FourierFT, # Params=480K2026.05 | 94.84 | |
| SOT-GLPShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 94.8 | |
| LoRABackbone=ViT-Large, Memory=1229.9 MB2026.05 | 94.8 | |
| MMRLmode=Target Evaluation2025.05 | 94.67 | |
| WCABackbone=L/14, Zero-shot=true2026.05 | 94.66 | |
| Uni-LoRABackbone=ViT-Large, Method=Uni-LoRA, # Params=144K2026.05 | 94.65 | |
| MMRL++mode=Target Evaluation2025.05 | 94.63 | |
| FourierFTBackbone=ViT-Large, Method=FourierFT, # Params=144K2026.05 | 94.46 | |
| CoOpOpmode=Target Evaluation2025.05 | 94.43 | |
| FFBackbone=ViT-Large, Method=FF, # Params=303.3M2026.05 | 94.43 | |
| FTBackbone=ViT-Large, Memory=4629.8 MB2026.05 | 94.4 | |
| EBMmode=Fine-tuning2025.10 | 94.4 | |
| CuPLBackbone=L/14, Zero-shot=true2026.05 | 94.33 | |
| GalLoPShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 94.1 | |
| Uni-LoRABackbone=ViT-Base, Method=Uni-LoRA, # Params=72K2026.05 | 94 | |
| TCPmode=Target Evaluation2025.05 | 93.97 | |
| GPartBackbone=ViT-Large, Method=GPart (Ours), # Params=144K2026.05 | 93.86 | |
| GPartBackbone=ViT-Base, Method=GPart (Ours), # Params=72K2026.05 | 93.85 | |
| MMAmode=Target Evaluation2025.05 | 93.8 | |
| SCANShot=162025.05 | 93.78 | |
| CoOpmode=Target Evaluation2025.05 | 93.7 | |
| PromptSRCShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 93.7 | |
| PromptSRCmode=Fine-tuning2025.10 | 93.7 | |
| CLIP-EBackbone=L/14, Zero-shot=true2026.05 | 93.62 | |
| PromptSRCmode=Target Evaluation2025.05 | 93.6 | |
| PLOTShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 93.6 | |
| PLOTmode=Fine-tuning2025.10 | 93.6 | |
| MaPLemode=Target Evaluation2025.05 | 93.53 | |
| ProDAShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 93.5 | |
| TOGAShots=8, Venue=-2026.03 | 93.4 | |
| TOGAShots=16, Venue=-2026.03 | 93.4 | |
| SVFitBackbone=ViT-Large, Memory=1351.5 MB2026.05 | 93.4 | |
| CoCoOpShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 93.3 | |
| CLIP-DBackbone=L/14, Zero-shot=true2026.05 | 93.3 | |
| SVFTBackbone=ViT-Large, Memory=1355.8 MB2026.05 | 93.3 | |
| CoCoOpmode=Fine-tuning2025.10 | 93.3 | |
| CLIPBackbone=L/14, Zero-shot=true2026.05 | 93.24 | |
| FourierFTBackbone=ViT-Base, Method=FourierFT, # Params=72K2026.05 | 93.21 | |
| CERSABackbone=ViT-Base, α=0.95, β=0.952026.05 | 93.2 | |
| CoFT+mode=unsupervised fine-tuning2026.02 | 93.16 | |
| FFBackbone=ViT-Base, Method=FF, # Params=85.8M2026.05 | 93.14 | |
| FTBackbone=ViT-Base2026.05 | 93.1 | |
| LoRABackbone=ViT-Base, Rank=322026.05 | 93.1 | |
| FourierFTBackbone=ViT-Base, Method=FourierFT, # Params=239K2026.05 | 93.05 | |
| CS-AlignerBackbone=ViT-B/16, Number of Shots=16, Evaluation Protocol=fine-tuned2025.02 | 93 | |
| MaPLeShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 92.8 | |
| MaPLemode=Fine-tuning2025.10 | 92.8 | |
| MMRL + MetaTPTadaptation=MetaTPT2025.12 | 92.79 | |
| LAGOBackbone=B/16, Zero-shot=true2026.05 | 92.78 | |
| CLIP-Adapter (best α)Backbone=ViT-B/16, Shots=16, Blending ratio (alpha) selection strategy=Best residual ratio per dataset (oracle)2026.03 | 92.7 | |
| CLIP-AdapterBackbone=ViT-B/16, Shots=16, Blending ratio (alpha)=best2026.03 | 92.7 | |
| TOGAShots=4, Venue=-2026.03 | 92.7 | |
| CLIP-Adapter (α = 0.2)Backbone=ViT-B/16, Shots=16, Blending ratio (alpha) selection strategy=Fixed residual ratio (α = 0.2), selected on ImageNet2026.03 | 92.6 | |
| CLIP-AdapterBackbone=ViT-B/16, Shots=16, Blending ratio (alpha)=0.22026.03 | 92.6 | |
| CAPTrole=Target2026.03 | 92.53 | |
| LwEIBrole=Target2026.03 | 92.5 | |
| SVFTBackbone=ViT-Base2026.05 | 92.5 | |
| CoFTmode=unsupervised fine-tuning2026.02 | 92.4 | |
| LoCoOpShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 92.4 | |
| CLIP-MoETraining shots=16-shot, Source dataset=ImageNet, Training epochs=32026.03 | 92.3 | |
| SVFitBackbone=ViT-Base, Rank=7682026.05 | 92.3 | |
| SCANShot=82025.05 | 92.29 | |
| HOSO-AdapterBackbone=ViT-B/16, Shots=16, Blending ratio (alpha) selection strategy=Optimised on hold-one-shot-out cache2026.03 | 92.27 | |
| HOSO-AdapterBackbone=ViT-B/16, Shots=16, Number of runs=32026.03 | 92.27 | |
| WCABackbone=B/16, Zero-shot=true2026.05 | 92.23 | |
| TOGAShots=2, Venue=-2026.03 | 92.2 | |
| CLIP-AdapterBackbone=ViT-B/16, Number of Shots=16, Evaluation Protocol=fine-tuned2025.02 | 92.1 | |
| CoOpShots=16, Backbone=ViT-B/16, Pre-training=CLIP2026.03 | 91.9 | |
| CoOpmode=Fine-tuning2025.10 | 91.9 | |
| VaMPCross-dataset evaluation protocol=true, Zero-shot transfer=true2025.11 | 91.79 | |
| GRIPmode=unsupervised fine-tuning2026.02 | 91.69 | |
| BiFTABackbone=B/16, Zero-shot=true2026.05 | 91.67 | |
| CLIP-MoE + MoE-GRPOTraining shots=16-shot, Source dataset=ImageNet, Training epochs=32026.03 | 91.6 | |
| PromptKDrole=Target2026.03 | 91.59 | |
| WaffleBackbone=L/14, Zero-shot=true2026.05 | 91.55 | |
| MMRLCross-dataset evaluation protocol=true, Zero-shot transfer=true2025.11 | 91.43 | |
| MMRLvariant=standard2025.12 | 91.43 | |
| MMRLadaptation=none2025.12 | 91.43 | |
| MMRL++Backbone=ViT-B/16, Training Shots=16-shot2026.05 | 91.43 | |
| MMRL2026.05 | 91.43 | |
| MMRL + NeRP2026.05 | 91.43 | |
| MMRLBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 91.43 | |
| MAILBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 91.37 | |
| A3B2Backbone=ViT-B/16, Training Shots=16-shot2026.05 | 91.36 | |
| SCANShot=42025.05 | 91.33 | |
| MMRL + TPTadaptation=TPT2025.12 | 91.31 | |
| MMRLBackbone=ViT-B/16, Training Shots=16-shot2026.05 | 91.31 | |
| TCPCross-dataset evaluation protocol=true, Zero-shot transfer=true2025.11 | 91.25 | |
| TCPBackbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 91.25 | |
| MMRLvariant=†2025.12 | 91.23 | |
| TOGAShots=1, Venue=-2026.03 | 91.2 | |
| MAIL++Backbone=ViT-B/16, Training Shots=16-shot, Training Source=ImageNet2026.05 | 91.2 | |
| CuPLBackbone=B/16, Zero-shot=true2026.05 | 91.14 | |
| LPBackbone=ViT-Large, Method=LP, # Params=02026.05 | 91.11 | |
| CLIP-SVD2025.09 | 91.06 |