List your company
ResearchFormerly WizWand

Image Classification on EMNIST Balanced (test)

91.06AccuracyWaveMix-128/7
-1.18822.76146.7170.659Jan 13, 2020Jul 20, 2020Jan 26, 2021Aug 4, 2021Feb 10, 2022Aug 19, 2022Feb 25, 2023
Updated 7mo ago

Evaluation results

MethodLinks
WaveMix-128/7Number of parameters (Million)=2.42022.0391.06———
WaveMixAugmentation=TrivialAugment2022.0591.06———
Previous SOTA [73]2022.0591.05———
VGG-5 (Spinal FC)Size of Fully Connected Layer=4HL, 128 Neurons Per Layer, Epoch=200, Parameters=3.630M2020.0790.73—91.050.1
VGG-5Size of Fully Connected Layer=1HL, 512 Neurons, Epoch=200, Parameters=3.646M2020.0790.71—91.04—
WaveMix-256/7Number of parameters (Million)=9.62022.0390.36———
ResNet-50Number of parameters (Million)=23.62022.0389.76———
ResNet-34Number of parameters (Million)=21.32022.0389.17———
ResNet-18Number of parameters (Million)=11.22022.0389———
CNN (Spinal FC)Size of Fully Connected Layer=6HL, 10 Neurons Per Layer, Epoch=8, Parameters=16.05k2020.0782.57—83.2117.66
CentralizedArchitecture=MLP (2 hidden layers, 64 neurons)2022.1082.45———
CNN (Spinal FC)Size of Fully Connected Layer=6HL, 8 Neurons Per Layer, Epoch=8, Parameters=13.82k2020.0782.13—82.7715.5
CNNSize of Fully Connected Layer=1HL, 50 Neurons, Epoch=8, Parameters=21.84k2020.0778.99—79.61—
DReS-FLArchitecture=PINN (2 hidden layers, 64 neurons), degree of gradient=82022.1078.04———
FedAvg-ISArchitecture=MLP (2 hidden layers, 64 neurons)2022.1077.09———
FedCILLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0273.12———
FedAvgArchitecture=MLP (2 hidden layers, 64 neurons)2022.1071.5———
FedAvg+ACGAN ReplayLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0266.87———
FedProx+ACGAN ReplayLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0266.19———
MCLmodel type=flexible models, presence of unlabeled samples=false, #n=131600, #f=784, #c=472020.0165.41———
FedAvg+DGRLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0263.55———
FedProx+DGRLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0263.55———
SCAFFOLDArchitecture=MLP (2 hidden layers, 64 neurons)2022.1055.15———
Fwdmodel type=flexible models, presence of unlabeled samples=false, #n=131600, #f=784, #c=472020.0118.21———
FedProxLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0217.74———
FedAvgLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0217.25———
FedLwF-2TLocal training iteration (T)=400, Global communication round (R)=200, Mini-batch size (B)=32, Backbone=3-layer CNN, Optimizer=Adam, Learning rate=1e-42023.0217.22———
PCmodel type=flexible models, presence of unlabeled samples=false, #n=131600, #f=784, #c=472020.0114.28———
MCULmodel type=flexible models, presence of unlabeled samples=true, #n=131600, #f=784, #c=472020.016.77———
GAmodel type=flexible models, presence of unlabeled samples=false, #n=131600, #f=784, #c=472020.014.25———
MCLmodel type=flexible models, presence of unlabeled samples=true, #n=131600, #f=784, #c=472020.014.14———
Fwdmodel type=flexible models, presence of unlabeled samples=true, #n=131600, #f=784, #c=472020.012.68———
GAmodel type=flexible models, presence of unlabeled samples=true, #n=131600, #f=784, #c=472020.012.54———
PCmodel type=flexible models, presence of unlabeled samples=true, #n=131600, #f=784, #c=472020.012.36———
DenseNet100-12Activation setting=Base2018.11—8.81——
DenseNet100-12Activation setting=Drop-Activation2018.11—8.9——
PreResNet164Activation setting=Base2018.11—8.88——
PreResNet164Activation setting=Drop-Activation2018.11—8.72——
ResNet164Activation setting=Base2018.11—8.85——
ResNet164Activation setting=Drop-Activation2018.11—8.82——
ResNeXt29, 8*64Activation setting=Base2018.11—9.07——
ResNeXt29, 8*64Activation setting=Drop-Activation2018.11—8.91——
WRN28-10Activation setting=Base2018.11—8.97——
WRN28-10Activation setting=Drop-Activation2018.11—8.72——