List your company
ResearchFormerly WizWand

Image Classification on pixel-by-pixel MNIST (test)

99.4AccuracyLSTM w/f STAR
6.611230.700654.7978.8794Feb 26, 2019Dec 12, 2019Sep 26, 2020Jul 12, 2021Apr 27, 2022Feb 10, 2023Nov 26, 2023
Updated 8mo ago

Evaluation results

MethodLinks
LSTM w/f STARNumber of layers=8, units=1282019.1199.4
STARNumber of layers=8, units=1282019.1199.2
STARNumber of layers=12, units=1282019.1199.2
IndRNNunits=1282019.1199
BN-LSTMunits=1002019.1199
AntisymmetricRNN w/ gating# units=128, # params=10k2019.0298.8
GRUNumber of layers=2, units=1282019.1198.8
AntisymRNNunits=1282019.1198.8
LSTMNumber of layers=2, units=1282019.1198.4
RHNNumber of layers=2, units=1282019.1198.4
STANH-RNNunits=1282019.1198.1
AntisymmetricRNN# units=128, # params=10k2019.0298
LSTM# units=128, # params=68k2019.0297.3
iRNNunits=1002019.1197
FC uRNN# units=512, # params=270k2019.0296.9
FC URNNunits=5122019.1196.9
KRU# units=512, # params=11k2019.0296.4
uRNNunits=5122019.1195.1
Soft orthogonal# units=128, # params=18k2019.0294.1
Soft orthounits=1282019.1194.1
FC uRNN# units=116, # params=16k2019.0292.8
Ours+LTraining Dataset=TextureMNIST2023.1161.29
Ours+LTraining Dataset=ColorMNIST2023.1138.88
CDEPTraining Dataset=ColorMNIST2023.1133.11
vRNNNumber of layers=1, units=1282019.1124.3
BaseTraining Dataset=ColorMNIST2023.1115.87
BaseTraining Dataset=TextureMNIST2023.1111.35
CDEPTraining Dataset=TextureMNIST2023.1110.18