List your company
ResearchFormerly WizWand

Image Classification on Food (Accuracy)

94.9AccuracyCLIP* L/14
19.177638.836358.49578.1537Jun 13, 2023Dec 11, 2023Jun 9, 2024Dec 7, 2024Jun 6, 2025Dec 4, 2025Jun 3, 2026
Updated 5mo ago

Evaluation results

MethodLinks
CLIP* L/14MAP head=true, Grain=Fine, Frozen representation=true, backbone=ViT-L/142023.0694.9
CapPa L/14MAP head=true, Grain=Fine, Frozen representation=true, backbone=ViT-L/142023.0694.2
CLIP* (16k)MAP head=true, Grain=Fine, Frozen representation=true2023.0692.1
ZSModel=MLLM, Method=ZS2026.0191.85
CapPaMAP head=true, Grain=Fine, Frozen representation=true2023.0691.5
CapMAP head=true, Grain=Fine, Frozen representation=true2023.0691.1
CLIP* (8k)MAP head=true, Grain=Fine, Frozen representation=true2023.0691
GPUASamples per class=full training set2026.0689.5
MMFTModel=ViT, Method=Ours2026.0188.79
GPUA*Samples per class=162026.0687.9
FLYPModel=ViT, Method=FLYP2026.0187.88
MMFTModel=RN50, Method=Ours2026.0187.61
CoCoOpShot=162026.0587.4
FTModel=ViT, Method=FT2026.0187.23
KgCoOpShot=162026.0587.2
CLIP-AdapterShot=162026.0587.1
PLOT++Shot=162026.0587.1
StatA2026.0687.1
FLYPModel=RN50, Method=FLYP2026.0186.93
KgCoOpShot=42026.0586.9
ZLaP2026.0686.9
TIP-AdapterShot=162026.0586.8
ZERO2026.0686.8
COSMICInference-time adaptation=true2026.0686.6
TIP-AdapterShot=42026.0586.5
CLIP-AdapterShot=42026.0586.5
PLOT++Shot=42026.0586.5
PARA2026.0486.4
KgCoOpShot=12026.0586.4
FTModel=RN50, Method=FT2026.0186.32
CoCoOpShot=42026.0586.3
PLOT++Shot=12026.0586.2
DPE2026.0686.2
CLIPShot=02026.0586.1
CLIP-AdapterShot=12026.0586.1
TDA2026.0686.1
CLIP ViT-B/16Img. encoder=CLIP ViT-B/162026.0586.04
CoCoOpFMA=false2025.1086
TIPPLEInference-time adaptation=true2026.0686
CLIPFMA=false2025.1085.9
CLIP2026.0685.9
TIP-AdapterShot=12026.0585.8
ProGradShot=162026.0585.8
+DP-FMShot=162026.0585.8
ProGradShot=42026.0585.4
CLIPFMA=true2025.1085.2
CoCoOpFMA=true2025.1085.2
CLIP-LoRAShot=12026.0585.1
DMN2026.0685.1
+DP-FMShot=12026.0585
MTA2026.0685
CoCoOpShot=12026.0584.9
ProGradShot=12026.0584.9
LoRAFMA=false2025.1084.5
CoOpShot=42026.0584.5
CoOpShot=162026.0584.4
CoOpShot=12026.0584.3
AdapterFMA=false2025.1084.2
CLIP-LoRAShot=162026.0584.2
+DP-FMShot=42026.0583.7
LoRAFMA=true2025.1083.6
AdapterFMA=true2025.1083.3
CLIP-LoRAShot=42026.0582.6
CoOpFMA=false2025.1080.8
GoRA2026.0480.66
CLIP ViT-B/32Img. encoder=CLIP ViT-B/322026.0580.53
CoOpFMA=true2025.1080.2
CLIP RN-101Img. encoder=CLIP RN-1012026.0579.83
LoRA2026.0479.43
LGCLIPBackbone=ViT-L/14, Correction (Manual Calibration)=false2025.1278.98
ZSModel=ViT, Method=ZS2026.0177.66
ZSModel=RN50, Method=ZS2026.0176.67
CLIP RN-50Img. encoder=CLIP RN-502026.0575.99
CLIPBackbone=ViT-L/14, Correction (Manual Calibration)=false2025.1275.4
DoRA2026.0474.49
SoRA2026.0472.19
LGCLIPBackbone=ViT-B/16, Correction (Manual Calibration)=false2025.1271.13
CLIPBackbone=ViT-B/16, Correction (Manual Calibration)=false2025.1269.97
AdaLoRA2026.0468.43
LGCLIPBackbone=ViT-B/32, Correction (Manual Calibration)=false2025.1263.17
ConvNext-B + CLIPImg. encoder=ConvNext-B, Txt. encoder=CLIP2026.0560.29
CLIPBackbone=ViT-B/32, Correction (Manual Calibration)=false2025.1260.15
CAFormer-S18 + CLIPImg. encoder=CAFormer-S18, Txt. encoder=CLIP2026.0559.83
TinyViT-21M + CLIPImg. encoder=TinyViT-21M, Txt. encoder=CLIP2026.0557.3
BEiT-B/16 + CLIPImg. encoder=BEiT-B/16, Txt. encoder=CLIP2026.0557.12
ConvFormer-S18 + CLIPImg. encoder=ConvFormer-S18, Txt. encoder=CLIP2026.0554.92
LGCLIPBackbone=RN50, Correction (Manual Calibration)=false2025.1252.61
CLIPBackbone=RN50, Correction (Manual Calibration)=false2025.1251.12
CAFormer-S18 + RoBERTaImg. encoder=CAFormer-S18, Txt. encoder=RoBERTa2026.0534.54
CAFormer-S18 + MPNetImg. encoder=CAFormer-S18, Txt. encoder=MPNet2026.0532.69
ConvFormer-S18 + RoBERTaImg. encoder=ConvFormer-S18, Txt. encoder=RoBERTa2026.0532.3
ConvFormer-S18 + MPNetImg. encoder=ConvFormer-S18, Txt. encoder=MPNet2026.0530.61
ConvNext-B + MPNetImg. encoder=ConvNext-B, Txt. encoder=MPNet2026.0530.52
BEiT-B/16 + RoBERTaImg. encoder=BEiT-B/16, Txt. encoder=RoBERTa2026.0529.48
ConvNext-B + RoBERTaImg. encoder=ConvNext-B, Txt. encoder=RoBERTa2026.0528.75
ConvNext-B + MiniLMImg. encoder=ConvNext-B, Txt. encoder=MiniLM2026.0527.55
BEiT-B/16 + MPNetImg. encoder=BEiT-B/16, Txt. encoder=MPNet2026.0527.2
TinyViT-21M + MPNetImg. encoder=TinyViT-21M, Txt. encoder=MPNet2026.0526.59
BEiT-B/16 + MiniLMImg. encoder=BEiT-B/16, Txt. encoder=MiniLM2026.0525.24
TinyViT-21M + RoBERTaImg. encoder=TinyViT-21M, Txt. encoder=RoBERTa2026.0522.09