List your company
ResearchFormerly WizWand

Incremental Learning on Tiny-ImageNet 5 task

55.25Average Incremental AccuracyUpper bound
25.058832.896940.73548.5731May 12, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Upper boundModel Inversion=false, Backbone=ResNet-32, Number of runs=52026.0555.25
Upper boundModel inversion=false, Backbone=ResNet-32, Number of runs=52026.0555.25
REMIXModel Inversion=true, Backbone=ResNet-32, Number of runs=52026.0538.64
REMIXModel inversion=true, Backbone=ResNet-32, Number of runs=52026.0538.64
PMIModel Inversion=true, Backbone=ResNet-32, Number of runs=52026.0537.9
PMIModel inversion=true, Backbone=ResNet-32, Number of runs=52026.0537.9
PMI w/o CFSModel Inversion=true, Backbone=ResNet-32, Number of runs=52026.0537.65
PMI w/o CFSModel inversion=true, Backbone=ResNet-32, Number of runs=52026.0537.65
DCMIModel inversion=true, Backbone=ResNet-32, Number of runs=52026.0535.78
R-DFCILModel Inversion=true, Backbone=ResNet-32, Number of runs=52026.0535.33
R-DFCILModel inversion=true, Backbone=ResNet-32, Number of runs=52026.0535.33
PRAKAModel inversion=false, Backbone=ResNet-32, Number of runs=52026.0531.69
ABDModel Inversion=true, Backbone=ResNet-32, Number of runs=52026.0530.83
ABDModel inversion=true, Backbone=ResNet-32, Number of runs=52026.0530.83
SSREModel inversion=false, Backbone=ResNet-32, Number of runs=52026.0526.22