List your company
ResearchFormerly WizWand

Image Classification on SUN397

98.4AccuracyLaCLIP
70.455277.710184.96592.2199Jun 17, 2021Mar 27, 2022Jan 4, 2023Oct 14, 2023Jul 23, 2024May 2, 2025Feb 10, 2026
Updated 8mo ago

Evaluation results

MethodLinks
LaCLIPArchitecture=ViT-B/16, Pre-training Data=LAION-400M, Protocol=5-way 5-shot2023.0598.4——
LaCLIPArchitecture=ViT-B/32, Pre-training Data=LAION-400M, Protocol=5-way 5-shot2023.0598.2——
CLIPArchitecture=ViT-B/16, Pre-training Data=LAION-400M, Protocol=5-way 5-shot2023.0598.2——
LaSLIPArchitecture=ViT-B/16, Pre-training Data=CC12M, Protocol=5-way 5-shot2023.0598.1——
CLIPArchitecture=ViT-B/32, Pre-training Data=LAION-400M, Protocol=5-way 5-shot2023.0598——
LaCLIPArchitecture=ViT-B/16, Pre-training Data=CC12M, Protocol=5-way 5-shot2023.0598——
SLIPArchitecture=ViT-B/16, Pre-training Data=CC12M, Protocol=5-way 5-shot2023.0597.5——
CLIPArchitecture=ViT-B/16, Pre-training Data=CC12M, Protocol=5-way 5-shot2023.0597.2——
LaCLIPArchitecture=ViT-B/16, Pre-training Data=RedCaps, Protocol=5-way 5-shot2023.0596.2——
CLIPArchitecture=ViT-B/16, Pre-training Data=RedCaps, Protocol=5-way 5-shot2023.0595.9——
LaCLIPArchitecture=ViT-B/16, Pre-training Data=CC3M, Protocol=5-way 5-shot2023.0595.2——
CLIPArchitecture=ViT-B/16, Pre-training Data=CC3M, Protocol=5-way 5-shot2023.0594.1——
TPTEvaluation Protocol=Cross-dataset transfer evaluation2024.0284.67——
X-FM_baseLinear evaluation=true, Model size=Base, Patch size=16*16, Resolution=224*2242023.0183.9——
fine-tunedBackbone=ViT-L/142026.0282.8——
Multiple modelsBackbone=ViT-L/142022.0882.4——
FLAVALinear evaluation=true, Model size=Base, Patch size=16*16, Resolution=224*2242023.0182.1——
MLCDPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0782.1——
CLIPPre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0781.8——
CLIP+Pre-training Data=WIT-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0781.5——
ISO-CTS/LARVBackbone=ViT-L/142026.0281.4—0.3
OpenCLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0781——
Joint patchingBackbone=ViT-L/142022.0880.9——
ISO-C/LARVBackbone=ViT-L/142026.0280.8—0.1
2SFS LoRABackbone=ViT-L/14, Shots (k=16)=16, Evaluation Protocol=all-to-all2025.0380.7——
Seq. patchingBackbone=ViT-L/14, Seed=02022.0880——
UNICOMPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Linear Probe2024.0780——
PyramidCLIPImage Encoder=ViT-B/16, Pretrain Dataset Size=143M, Evaluation Protocol=Linear probe2022.0479.9——
TSV-M/LARVBackbone=ViT-L/142026.0279.8—1.6
Multiple modelsBackbone=ViT-B/162022.0879——
IndividualBackbone=ViT-B/16, Pre-trained=ImageNet-21k, Number of Merged Models=12024.0578.98——
fine-tunedBackbone=ViT-B/162026.0278.9——
Seq. patchingBackbone=ViT-L/14, Seed=12022.0878.8——
CLIP*Image Encoder=ViT-B/16, Pretrain Dataset Size=400M, Evaluation Protocol=Linear probe2022.0478.4——
Seq. patchingBackbone=ViT-L/14, Seed=22022.0878.4——
CLIPLinear evaluation=true, Model size=Base, Patch size=16*16, Resolution=224*2242023.0178.4——
DaVinciLinear evaluation=true, Model size=Base, Patch size=16*16, Resolution=224*2242023.0178——
MMRLShots=16 shots2025.0377.7——
AWTShots=162024.0777.57——
PyramidCLIPImage Encoder=ViT-B/32, Pretrain Dataset Size=143M, Evaluation Protocol=Linear probe2022.0477.5——
TIES/LARVBackbone=ViT-L/142026.0277.3—2.6
ProSRCShot=16, Time (s)=14403, Mem. (M)=33732024.1277.23——
PromptSRCShots=16 shots2025.0377.23——
2SFS LoRABackbone=ViT-B/16, Shots (k=16)=16, Evaluation Protocol=all-to-all2025.0376.9——
Skip TuningShot=16, Time (s)=962, Mem. (M)=5282024.1276.8——
ISO-C/LARVBackbone=ViT-B/162026.0276.7—0.9
CLIP*Image Encoder=ViT-B/32, Pretrain Dataset Size=400M, Evaluation Protocol=Linear probe2022.0476.6——
BASIC-LBackbone=BASIC-L2021.1176.2——
EMR-MERGINGBackbone=ViT-B/16, Pre-trained=ImageNet-21k, Number of Merged Models=302024.0576.19——
PyramidCLIPTask=Linear Probe, Pretrain Dataset=143M, Image Encoder=ResNet502022.0476.1——
MMRLShots=8 shots2025.0376——
Joint patchingBackbone=ViT-B/162022.0875.9——
AdaMergingBackbone=ViT-L/142026.0275.9——
ProSRCShot=8, Time (s)=7636, Mem. (M)=33732024.1275.73——
PromptSRCShots=8 shots2025.0375.73——
MaPLeShot=16, Time (s)=14070, Mem. (M)=30612024.1275.53——
MaPLeShots=16 shots2025.0375.53——
ISO-CTS/LARVBackbone=ViT-B/162026.0275.5—-0.6
Multiple modelsBackbone=ViT-B/322022.0875.1——
fine-tunedBackbone=ViT-B/322026.0274.9——
Skip TuningShot=8, Time (s)=508, Mem. (M)=5282024.1274.77——
CLIPTask=Linear Probe, Pretrain Dataset=143M, Image Encoder=ResNet502022.0474.7——
CoOpShot=16, Time (s)=3879, Mem. (M)=45512024.1274.67——
CoOpShots=16 shots2025.0374.67——
MMAShots=16 shots2025.0374.63——
2SFS LoRABackbone=ViT-B/32, Shots (k=16)=16, Evaluation Protocol=all-to-all2025.0374.5——
TA/LARVBackbone=ViT-L/142026.0274.5—2.5
ProSRCShot=4, Time (s)=3841, Mem. (M)=33732024.1274——
PromptSRCShots=4 shots2025.0374——
TSV-M/LARVBackbone=ViT-B/162026.0274—0.9
MMRLShots=4 shots2025.0373.93——
Seq. patchingBackbone=ViT-B/16, Seed=02022.0873.4——
Parallel patchingBackbone=ViT-L/142022.0873.4——
CLIPTask=Linear Probe, Pretrain Dataset=400M, Image Encoder=ResNet502022.0473.3——
CLIPBackbone=ResNet-50, Evaluation Protocol=Linear probe2021.0673.3——
Linear probe CLIPShots=16 shots2025.0373.28——
MaPLeShot=8, Time (s)=7060, Mem. (M)=30612024.1273.23——
MaPLeShots=8 shots2025.0373.23——
FLIPPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0773.1——
Skip TuningShot=4, Time (s)=280, Mem. (M)=5282024.1273.07——
BASIC-MBackbone=BASIC-M2021.1172.9——
ISO-CTS/LARVBackbone=ViT-B/322026.0272.9—1.8
DeCLIPTask=Linear Probe, Pretrain Dataset=88M, Image Encoder=ResNet502022.0472.8——
StatAEncoder=ViT-L/14, K_eff=All2025.0172.8——
AWTShots=12024.0772.65——
Joint patchingBackbone=ViT-B/322022.0872.6——
Simple AveragingBackbone=ViT-L/142026.0272.5——
Seq. patchingBackbone=ViT-B/16, Seed=12022.0872.4——
MMAShots=8 shots2025.0372.3——
TIES/LARVBackbone=ViT-B/162026.0272.2—1.5
CoCoOpShot=16, Time (s)=10694, Mem. (M)=28882024.1272.15——
CoCoOpShots=16 shots2025.0372.15——
MLCDPre-training Data=LAION-400M, Backbone=ViT-L/14, Evaluation Protocol=Zero-shot classification2024.0772.1——
StatABackbone=ViT-L/14, K_eff=Very High, Batch Size=1000, Evaluation Protocol=batch test-time adaptation2025.0171.6——
StatABackbone=ViT-L/14, Keff=Low2025.0171.6——
ProSRCShot=2, Time (s)=1947, Mem. (M)=33732024.1271.6——
PromptSRCShots=2 shots2025.0371.6——
CoOpShot=8, Time (s)=1988, Mem. (M)=45512024.1271.53——
CoOpShots=8 shots2025.0371.53——
MMRLShots=2 shots2025.0371.53——