List your company

Face Verification on Labeled Faces in the Wild (LFW)

99.17AccuracyAM-Softmax
96.996497.560798.12598.6893Jan 17, 2018

Evaluation results

MethodLinks
AM-Softmaxm=0.4, Backbone=ResNet-20, Feature Normalization=true2018.0199.17————
AM-Softmaxm=0.25, Backbone=ResNet-20, Feature Normalization=true2018.0199.13————
AM-Softmax w/o FNm=0.4, Backbone=ResNet-20, Feature Normalization=false2018.0199.12————
AM-Softmaxm=0.5, Backbone=ResNet-20, Feature Normalization=true2018.0199.1————
A-Softmaxm=~1.5, Backbone=ResNet-20, Feature Normalization=true2018.0199.08————
AM-Softmaxm=0.3, Backbone=ResNet-20, Feature Normalization=true2018.0199.08————
AM-Softmax w/o FNm=0.35, Backbone=ResNet-20, Feature Normalization=false2018.0199.08————
AM-Softmaxm=0.45, Backbone=ResNet-20, Feature Normalization=true2018.0199.03————
Center LossBackbone=ResNet-28, Feature Normalization=false2018.0199————
NormFaceBackbone=ResNet-28, Feature Normalization=true, Fine-tuned on Center Loss=true2018.0198.98————
AM-Softmaxm=0.35, Backbone=ResNet-20, Feature Normalization=true2018.0198.98————
Softmax + 75% dropoutBackbone=ResNet-20, Feature Normalization=true, Dropout=75%2018.0198.62————
SoftmaxBackbone=ResNet-20, Feature Normalization=true2018.0197.08————
3DDFA3D=true, Texture=false2016.12————10
AugNet2016.12———99.594.2
DeepFace-ensemble2016.12———99.693.7
EigenFaces2016.12———256.2
Hybrid Descriptor2016.12———66.642.4