List your company
ResearchFormerly WizWand

Image Classification on Describable Textures Dataset (DTD) (test)

78Accuracy2SFS
40.476850.218459.9669.7016Nov 3, 2020Oct 14, 2021Sep 24, 2022Sep 5, 2023Aug 15, 2024Jul 26, 2025Jul 7, 2026
Updated 3mo ago

Evaluation results

MethodLinks
2SFSBackbone=ViT-L/14, Shots=162025.0378
SCEProtocol=Linear classification2022.1277.1
MoCo V2 Anisotropictraining_mode=SSL, texture_suppression=anisotropic2020.1176.3
BYOLtraining_mode=SSL2020.1176.2
SimCLRProtocol=Linear classification2022.1275.7
BYOLProtocol=Linear classification2022.1275.5
NNCLRProtocol=Linear classification2022.1275.5
SupervisedProtocol=Linear classification2022.1274.9
MoCo V2training_mode=SSL2020.1174.7
SimCLRtraining_mode=SSL2020.1173.2
2SFSBackbone=ViT-B/16, Shots=162025.0373.2
Supervised (Reproduced)training_mode=supervised2020.1172.66
2SFSBackbone=ViT-B/32, Shots=162025.0370.4
SHiRABackbone=ViT-Base, Evaluation Protocol=Fine-tuned2024.0669.73
LoRABackbone=ViT-Base, Evaluation Protocol=Fine-tuned2024.0669.41
TCP + SAMPLeOptimizer=SAMPLe2026.0747.79
ALIPEvaluation Protocol=Linear probing2026.0547.1
CoPromptOptimizer=Baseline2026.0747.07
DINORANKCLIP (full)Evaluation Protocol=Linear probing2026.0547
CoCoOp + SAMPLeOptimizer=SAMPLe2026.0746.75
MaPLeOptimizer=Baseline2026.0746.49
CyCLIPEvaluation Protocol=Linear probing2026.0545.8
CoCoOpOptimizer=Baseline2026.0745.73
CoOp + SAMPLeOptimizer=SAMPLe2026.0744.86
MaPLe + SAMPLeOptimizer=SAMPLe2026.0744.55
TCPOptimizer=Baseline2026.0744.35
CLIPEvaluation Protocol=Linear probing2026.0543.2
Ours w/o DINOEvaluation Protocol=Linear probing2026.0543
CoPrompt + SAMPLeOptimizer=SAMPLe2026.0742.74
RANKCLIPEvaluation Protocol=Linear probing2026.0542.4
CoOpOptimizer=Baseline2026.0741.92