List your company
ResearchFormerly WizWand

Image Classification on SAT-4 (test)

99.9AccuracyDeepSat V2
79.183284.561689.9495.3184Sep 11, 2015May 22, 2016Jan 31, 2017Oct 13, 2017Jun 24, 2018Mar 5, 2019Nov 15, 2019
Updated 8mo ago

Evaluation results

MethodLinks
DeepSat V22019.1199.9
Triplet networks2019.1199.76
SatCNNNormalization=Z-score2019.1199.69
SatCNNNormalization=linear2019.1199.55
D-DSML-CaffeNet2019.1199.51
Contrastive loss2019.1198.74
TradCNNNormalization=Z-score2019.1198.43
DCNN2019.1198.41
DeepSat2019.1197.95
DeepSatNeurons per layer=50, Layers=22015.0997.946
DeepSatNeurons per layer=50, Layers=32015.0997.654
DeepSatNeurons per layer=100, Layers=22015.0997.292
DeepSatNeurons per layer=20, Layers=22015.0997.115
DeepSatNeurons per layer=10, Layers=32015.0996.8
DeepSatNeurons per layer=10, Layers=22015.0996.585
DeepSatNeurons per layer=100, Layers=32015.0995.609
DeepSatNeurons per layer=20, Layers=32015.0995.473
MLPNormalization=Z-score2019.1194.76
CNN2019.1186.83
DBN2019.1181.78
SDAE2019.1179.98