Image Classification on ImageNet-Sketch (Acc, NLPD, ECE)
72.9AccuracyXray
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| XrayViT=EViT-2B, Backbone Type=Ours, Resolution=336, Active Tokens=2882026.02 | 72.9 | — | — | |
| PEcoreViT=G/14, Backbone Type=Weakly-supervised backbones2026.02 | 71.9 | — | — | |
| SigLIP 2ViT=g/16, Backbone Type=Weakly-supervised backbones2026.02 | 71.8 | — | — | |
| DINOv3ViT=7B/16, Backbone Type=Self-supervised backbones2026.02 | 71.3 | — | — | |
| AM-RADIOv2.5ViT=g/14, Backbone Type=Agglomerative backbones2026.02 | 67.1 | — | — | |
| AIMv2ViT=3B/14, Backbone Type=Weakly-supervised backbones2026.02 | 67.1 | — | — | |
| DINOv2ViT=g/14, Backbone Type=Self-supervised backbones2026.02 | 65.4 | — | — | |
| EVA-CLIPViT=18B/14, Backbone Type=Weakly-supervised backbones2026.02 | 64 | — | — | |
| Web-DINOViT=7B/14, Backbone Type=Self-supervised backbones2026.02 | 64 | — | — | |
| ViT-B/16Architecture=ViT-B/16, Zero-shot=true2026.03 | 52.4 | — | — | |
| CoOpOp2026.05 | 51.62 | — | — | |
| RPO2026.05 | 50.67 | — | — | |
| CAPTShot=16, Backbone=ViT-B/16, Domain Role=Target2026.03 | 50.46 | — | — | |
| SeMoBridge-TBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training2025.09 | 50.44 | — | — | |
| TACShot=16, Backbone=ViT-B/16, Domain Role=Target2026.03 | 50.3 | — | — | |
| LwEIBShot=16, Backbone=ViT-B/16, Domain Role=Target2026.03 | 50.07 | — | — | |
| APEBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training-free2025.09 | 49.95 | — | — | |
| APE-TBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training2025.09 | 49.73 | — | — | |
| SeMoBridgeBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training-free2025.09 | 49.55 | — | — | |
| FrancaViT=g/14, Backbone Type=Self-supervised backbones2026.02 | 49.5 | — | — | |
| DOCOSetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 49.5 | — | — | |
| E-COMESetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 49.2 | — | — | |
| MaPLeBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training2025.09 | 49.15 | — | — | |
| DPCoreSetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 49 | — | — | |
| CLIP-Adapter2026.05 | 48.94 | — | — | |
| LDCBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training2025.09 | 48.85 | — | — | |
| CoCoOpBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training2025.09 | 48.75 | — | — | |
| T3AGF=true, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 48.44 | — | — | |
| CoOpShot=16, Backbone=ViT-B/16, Domain Role=Target2026.03 | 47.99 | — | — | |
| CoOpBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Training2025.09 | 47.99 | — | — | |
| ELaTTAGF=true, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 47.73 | — | — | |
| EATASetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 47.6 | — | — | |
| UniEntSetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 47.6 | — | — | |
| NEOBackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 47.2 | — | — | |
| SurgeonBackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 47 | — | — | |
| CoTTASetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 46.7 | — | — | |
| STAMPSetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 46.7 | — | — | |
| SourceSetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 46.6 | — | — | |
| TentSetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 46.6 | — | — | |
| OSTTASetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 46.6 | — | — | |
| FOABackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 46.3 | — | — | |
| CLIPBackbone=ViT-B/16, Shots=16, Evaluation Protocol=Zero-Shot2025.09 | 46.15 | — | — | |
| MEMOGF=false, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 46.08 | — | — | |
| PASLEGF=false, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 45.72 | — | — | |
| TENTBackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 45.7 | — | — | |
| ViDASetting=Closed-set, Scenario=Continual Test-Time Adaptation2026.04 | 45.6 | — | — | |
| T3ABackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 45.5 | — | — | |
| SARBackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 45.5 | — | — | |
| No AdaptBackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 45.4 | — | — | |
| CoTTABackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 45.2 | — | — | |
| LAMEBackbone=ViT-Base, Evaluation Protocol=Test-Time Adaptation2025.10 | 45 | — | — | |
| FOAGF=true, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 44.95 | — | — | |
| No AdaptGF=true, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 44.89 | — | — | |
| SARGF=false, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 44.82 | — | — | |
| CoTTAGF=false, Backbone=ViT-Base, Protocol=single-instance episodic2025.10 | 44.77 | — | — | |
| CLIP-RDArchitecture=ResNet-50, Zero-shot=true, Distillation Method=CLIP-RD2026.03 | 42.3 | — | — | |
| CLIP-KDArchitecture=ResNet-50, Zero-shot=true, Distillation Method=CLIP-KD2026.03 | 41.6 | — | — | |
| CLIP-RDArchitecture=EfficientNet-B0, Zero-shot=true, Distillation Method=CLIP-RD2026.03 | 33.3 | — | — | |
| CLIP-RDArchitecture=ResNet-18, Zero-shot=true, Distillation Method=CLIP-RD2026.03 | 26.2 | — | — | |
| CLIP-KDArchitecture=ResNet-18, Zero-shot=true, Distillation Method=CLIP-KD2026.03 | 25.8 | — | — | |
| ResNet-50Architecture=ResNet-50, Zero-shot=true2026.03 | 24.4 | — | — | |
| ResNet-18Architecture=ResNet-18, Zero-shot=true2026.03 | 18 | — | — | |
| EfficientNet-B0Architecture=EfficientNet-B0, Zero-shot=true2026.03 | 18 | — | — | |
| BayesVLMzero-shot=true2024.12 | — | — | 0.08 | |
| CaRotBackbone=ViT-B/162026.05 | — | — | 0.07 | |
| CLIPzero-shot=true2024.12 | — | — | 0.16 | |
| CLIPzero-shot=true, temperature scaling=true2024.12 | — | — | 0.13 | |
| FLYPBackbone=ViT-B/162026.05 | — | — | 0.22 | |
| Lipsum-FTBackbone=ViT-B/162026.05 | — | — | 0.102 | |
| LP-FTBackbone=ViT-B/16, Mode=Linear-probing followed by Fine-tuning2026.05 | — | — | 0.166 | |
| TRACERBackbone=ViT-B/162026.05 | — | — | 0.078 | |
| TTAzero-shot=true2024.12 | — | — | 0.15 | |
| ZSBackbone=ViT-B/16, Mode=Zero-shot2026.05 | — | — | 0.085 |