List your company
ResearchFormerly WizWand

Face Identification on MegaFace 1M distractors 1.0 (test)

96.2Rank-1 AccuracyDRConv-MobileFaceNet
31.205248.0788564.952581.82615Apr 26, 2017Oct 20, 2017Apr 16, 2018Oct 11, 2018Apr 6, 2019Oct 1, 2019Mar 27, 2020
Updated 8mo ago

Evaluation results

MethodLinks
DRConv-MobileFaceNetMADDs (x10^6)=2012020.0396.2
Local-MobileFaceNetMADDs (x10^6)=1892020.0394.9
CondConv-MobileFaceNetMADDs (x10^6)=1952020.0394.8
Vocord - deepVo V3Protocol=Large2018.0191.76
MobileFaceNetMADDs (x10^6)=1892020.0391.3
CosFace(3-patch ensemble)Protocol=Large2018.0184.26
YouTu LabProtocol=Large2018.0183.29
CosFace(Single-patch)Protocol=Large2018.0182.72
DeepSense V2Protocol=Large2018.0181.29
CosFace(3-patch ensemble)Protocol=Small2018.0179.54
FUDAN-CS_SDSProtocol=Small2018.0177.98
GRCCVProtocol=Small2018.0177.67
CosFace(Single-patch)Protocol=Small2018.0177.11
Beijing FaceAll V2Protocol=Small2018.0176.66
SphereFace (3-patch ensemble)protocol=Small2017.0475.766
SphereFace - SmallProtocol=Small2018.0175.76
Vocord - Deep Volprotocol=Large2017.0475.127
Deepsense - Largeprotocol=Large2017.0474.799
SIATMMLAB Tencent VisionProtocol=Large2018.0174.2
Shanghai Techprotocol=Large2017.0474.049
NTechLAB - facenx largeprotocol=Large2017.0473.3
NTechLAB - facenx_largeProtocol=Large2018.0173.3
SphereFace (single model)protocol=Small2017.0472.729
Deepsense - Smallprotocol=Small2017.0470.983
DeepSense - SmallProtocol=Small2018.0170.98
Google - FaceNet v8protocol=Large2017.0470.496
Google - FaceNet v8Protocol=Large2018.0170.49
L-Softmax Lossprotocol=Small2017.0467.128
Softmax+Center Lossprotocol=Small2017.0465.494
SIAT_MMLABprotocol=Small2017.0465.233
SIAT_MMLABProtocol=Small2018.0165.23
Softmax+Contrastive Lossprotocol=Small2017.0465.219
Beijing FaceAll_Norm_1600protocol=Large2017.0464.804
Beijing FaceAll_Norm_1600Protocol=Large2018.0164.8
Triplet Lossprotocol=Small2017.0464.797
Beijing FaceAll_1600protocol=Large2017.0463.977
Barebones FR - cnnprotocol=Small2017.0459.363
NTechLAB - facenx_smallprotocol=Small2017.0458.218
Softmax Lossprotocol=Small2017.0454.855
3DiVi Company - tdvm6protocol=Small2017.0433.705