List your company
ResearchFormerly WizWand

Image Classification on TissueMNIST

74.84AccuracyMAIL-RN18
12.408828.616944.82561.0331Feb 17, 2026Mar 5, 2026Mar 21, 2026Apr 6, 2026Apr 22, 2026May 8, 2026May 25, 2026
Updated 4mo ago

Evaluation results

MethodLinks
MAIL-RN18Base network=ResNet182026.0274.8496.56——
MAIL-M2Base network=MobileNetV22026.0274.1896.27——
DRIFA-Net2026.0273.7596.03——
MedVIT-S2026.0573.195.2——
MFMSA2026.0272.4595.29——
MO-MAE2026.057294.4——
M3Att2026.0270.5494.52——
AutoKeras2026.0570.394.1——
MedVIT-T2026.0570.394.3——
MedVIT-L2026.0569.993.5——
Gloria2026.0269.8894.72——
MuMu2026.0269.7594.19——
HAMLET2026.0269.5893.98——
MTTU-Net2026.0269.193.75——
DDA-Net2026.0268.994.54——
NAT2026.0268.4294.39——
ResNet-182026.0568.193.3——
ResNet-502026.056893.2——
POTTER2026.0267.9893.29——
Google AutoML2026.0567.392.4——
AMIMV-SSLBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0461.56———
MoCoV3Backbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0459.68———
DINOBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0459.39———
NNCLRBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0459.39———
SimCLRBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0458.15———
VICRegBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0456.2———
BYOLBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0455.59———
Barlow TwinsBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0453.89———
Auto-sklearn2026.0553.282.8——
ReSSLBackbone=ResNet-50, Initialization=Random, Training Strategy=Self-supervised2026.0448.45———
BioVLMTrain Params=30K2026.0427.34———
CoOpTrain Params=3K2026.0422.68———
TCPTrain Params=0.5M2026.0420.87———
KgCoOpTrain Params=3K2026.0420.81———
PromptSRCTrain Params=69K2026.0420.43———
ProGradTrain Params=3K2026.0418.76———
MaPLeTrain Params=5.3M2026.0418.35———
CoCoOpTrain Params=44K2026.0417.45———
BioMedCLIPTrain Params=02026.0415.7———
BiomedCoOpTrain Params=3K2026.0414.81———
AMIMV-SSLBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—89.77——
Barlow TwinsBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—84.16——
BYOLBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—86.11——
DINOBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—88.55——
EL2NRatio=2%2026.06———10.1
EL2NRatio=5%2026.06———15.2
EVARatio=2%2026.06———39.8
EVARatio=5%2026.06———46.7
FacilityRatio=2%2026.06———44.7
FacilityRatio=5%2026.06———48.1
Forg.Ratio=2%2026.06———39.3
Forg.Ratio=5%2026.06———43.6
FPSRatio=2%2026.06———35.6
FPSRatio=5%2026.06———41
Graph Coverage SelectionRatio=2%2026.06———42.6
Graph Coverage SelectionRatio=5%2026.06———49.5
HerdingRatio=2%2026.06———42.9
HerdingRatio=5%2026.06———48.3
Learnable PEPositional Encoding=Learnable2026.03——69.88—
MoCoV3Backbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—88.66——
NNCLRBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—88.41——
No PEPositional Encoding=None2026.03——69.79—
RandomRatio=2%2026.06———43.1
RandomRatio=5%2026.06———49.1
ResNet-18Backbone=ResNet-182026.03——69.88—
ReSSLBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—78.65——
RoPEPositional Encoding=RoPE2026.03——69.72—
SimCLRBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—87.38——
VICRegBackbone=ResNet-50, Initialization=random, Training Protocol=self-supervised2026.04—86.34——