List your company
ResearchFormerly WizWand

Multimodal Classification on DVM

99.17Multi Modality AccuracyGAAL
87.324490.399793.47596.5503Apr 2, 2026Apr 12, 2026Apr 22, 2026May 3, 2026May 13, 2026May 23, 2026Jun 3, 2026
Updated 4mo ago

Evaluation results

MethodLinks
GAALLearning Paradigm=Proposed Method2026.0499.1790.5791.91
DFPLMissing rate η=30%, Train/Test Image=70%, Train/Test Tabular=100%2026.0698.23——
DFPLMissing rate η=50%, Train/Test Image=50%, Train/Test Tabular=100%2026.0697.66——
DFPLMissing rate η=70%, Train/Test Image=30%, Train/Test Tabular=100%2026.0697.5——
LFMLearning Paradigm=Multimodal Gradient Conflict2026.0497.3185.4388.48
DI-MMLLearning Paradigm=Multimodal Gradient Conflict2026.0497.1984.9987.48
ReconBoostLearning Paradigm=Multimodal Gradient Conflict2026.0497.1484.9987.56
MLALearning Paradigm=Multimodal Gradient Conflict2026.0496.6876.6485.05
MMParetoLearning Paradigm=Multimodal Gradient Conflict2026.0496.5875.4988.51
DFPLMissing rate η=30%, Train/Test Image=85%, Train/Test Tabular=85%2026.0696.24——
DFPLMissing rate η=30%, Train/Test Image=100%, Train/Test Tabular=70%2026.0695.54——
TIPLearning Paradigm=Tabular-image Fusion2026.0495.45——
DAFTLearning Paradigm=Tabular-image Fusion2026.0494.6——
DFPLMissing rate η=50%, Train/Test Image=75%, Train/Test Tabular=75%2026.0694.5——
DFPLMissing rate η=50%, Train/Test Image=100%, Train/Test Tabular=50%2026.0693.43——
DFPLMissing rate η=70%, Train/Test Image=65%, Train/Test Tabular=65%2026.0692.63——
DFPLMissing rate η=70%, Train/Test Image=100%, Train/Test Tabular=30%2026.0692.59——
MFLearning Paradigm=Tabular-image Fusion2026.0489.93——
CFLearning Paradigm=Tabular-image Fusion2026.0487.930.5887.92
OGMLearning Paradigm=Multimodal Gradient Conflict2026.0487.780.7687.8
CHARMSLearning Paradigm=Test-time Tabular Missing2026.04—91.75—
FMRLearning Paradigm=Test-time Tabular Missing2026.04—84.27—
GAAL*Learning Paradigm=Test-time Tabular Missing, Pretrained weight setting=Same as test-time tabular missing methods2026.04—93.58—
KDLearning Paradigm=Test-time Tabular Missing2026.04—83.9—
MLPLearning Paradigm=Unimodal Learning2026.04——87.42
MMCLLearning Paradigm=Test-time Tabular Missing2026.04—82.03—
Resnet50Learning Paradigm=Unimodal Learning2026.04—87.43—