List your company
ResearchFormerly WizWand

Face Recognition on CFP-FP

99.51AccuracyPFC
82.547686.951391.35595.7587Sep 20, 2021Jul 2, 2022Apr 13, 2023Jan 24, 2024Nov 4, 2024Aug 16, 2025May 29, 2026
Updated 5mo ago

Evaluation results

MethodLinks
PFCTraining Dataset=WebFace42M, Backbone=ResNet-200, Sampling Ratio=0.32022.0399.51
ResNet200-TopoFRYear='24, Param (M)=118.8, FLOPs=23.5G, Training Dataset=Glint360K, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=12.242026.0499.45
PFCTraining Dataset=WebFace42M, Backbone=ViT-L, Sampling Ratio=0.32022.0399.44
PFCTraining Dataset=WebFace12M, Backbone=ResNet-100, Sampling Ratio=0.32022.0399.4
PFCTraining Dataset=WebFace42M, Backbone=ResNet-100, Sampling Ratio=0.32022.0399.4
PFCTraining Dataset=WebFace42M, Backbone=ViT-B, Sampling Ratio=0.32022.0399.4
Virtual FC2022.0399.38
TransFace-LYear='25, Param (M)=271.6, FLOPs=25.4G, Training Dataset=Glint360K, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=OOM2026.0499.37
BioMetricNet2022.0399.35
WebFace42M2022.0399.35
PFCTraining Dataset=WebFace42M, Backbone=ResNet-100, Sampling Ratio=0.0082022.0399.32
TransFace-BYear='25, Param (M)=124.5, FLOPs=11.5G, Training Dataset=Glint360K, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=18.202026.0499.24
PFCTraining Dataset=WebFace4M, Backbone=ResNet-100, Sampling Ratio=0.32022.0399.23
PFCTraining Dataset=WebFace12M, Backbone=ResNet-100, Sampling Ratio=0.0132022.0399.21
VPL2022.0399.11
PFCTraining Dataset=WebFace4M, Backbone=ResNet-100, Sampling Ratio=0.042022.0399.06
TransFace-SYear='25, Param (M)=86.7, FLOPs=5.8G, Training Dataset=Glint360K, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=14.312026.0498.96
Sub-center2022.0398.8
DUL2022.0398.78
ResNet50-ArcFaceYear='22, Param (M)=43.6, FLOPs=6.3G, Training Dataset=Glint360K, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=3.762026.0498.77
ElasticFace-Cos+Training Dataset=MS1MV2 [7], [4]2021.0998.73
ElasticFace-ArcTraining Dataset=MS1MV2 [7], [4]2021.0998.67
URFace2022.0398.64
GroupFaceTraining Dataset=clean MS1M [7], [4]2021.0998.63
GroupFace2022.0398.63
BroadFace2022.0398.63
ElasticFace-CosTraining Dataset=MS1MV2 [7], [4]2021.0998.61
ElasticFace-Arc+Training Dataset=MS1MV2 [7], [4]2021.0998.6
Partial-FC-CosFaceTraining Dataset=MS1MV2 [7], [4]2021.0998.51
VarGFaceNetYear='19, Param (M)=5.0, FLOPs=1.0G, Training Dataset=MS1MV3, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.832026.0498.5
MagFaceTraining Dataset=MS1MV2 [7], [4]2021.0998.46
Partial-FC-ArcFaceTraining Dataset=MS1MV2 [7], [4]2021.0998.45
DCQ2022.0398.44
CurricularFaceTraining Dataset=MS1MV2 [7], [4]2021.0998.37
CurricularFace2022.0398.37
KANFace-0.5Year='25, Param (M)=6.80, FLOPs=397M, Training Dataset=WebFace12M, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=9.982026.0498.31
ArcFaceVenue=CVPR '19, PPFR=No2024.0398.3
MV-Softmax2022.0398.28
ArcFaceTraining Dataset=MS1MV2 [7], [4]2021.0998.27
ArcFace2022.0398.27
FaceLiVTv2-LYear=-, Param (M)=8.52, FLOPs=309M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.71(14.0× ↓)2026.0498.26
CosFace2022.0398.12
FaceLiVTv2-MYear=-, Param (M)=7.02, FLOPs=258M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.65(16.7%↓)2026.0497.93
SwiftFaceFormer-L3Year='24, Param (M)=28.0, FLOPs=2.0G, Training Dataset=MS1MV3, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=2.592026.0497.8
DuetFaceVenue=MM '22, PPFR=Yes2024.0397.79
PartialFaceVenue=ICCV '23, PPFR=Yes2024.0397.63
FaceLiVTv2-SYear=-, Param (M)=4.62, FLOPs=179M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.54(30.8%↓)2026.0497.47
KANFace-0.5Year='25, Param (M)=6.80, FLOPs=397M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Retrained from scratch, Latency (ms) on iPhone 15 Pro=9.982026.0497.35
ShuffleFaceNet-1.5Year='21, Param (M)=2.6, FLOPs=577M, Training Dataset=MS1MV2, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.692026.0497.26
FaceLiVTv1-MYear='25, Param (M)=9.8, FLOPs=386M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.672026.0497.2
KANFace-0.6Year='25, Param (M)=4.74, FLOPs=240M, Training Dataset=WebFace12M, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=6.522026.0497.17
VirFace2022.0397.15
DCTDPVenue=ECCV '22, PPFR=Yes, Privacy Budget=epsilon=12024.0396.97
MinusFaceVenue=ours, PPFR=Yes, Backbone=IR-502024.0396.92
CASIAVenue=(Real), Generator=/, Scale (#IDs × #Imgs)=0.49M (10.5K × 47)2026.0596.91
MobileFaceNetYear='21, Param (M)=0.99, FLOPs=440M, Training Dataset=MS1MV2, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.772026.0496.9
GhostFaceNetV1-1Year='23, Param (M)=4.1, FLOPs=216M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.782026.0496.83
SwiftFaceFormer-L1Year='24, Param (M)=11.8, FLOPs=805M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=1.202026.0496.61
SwiftFaceFormer-SYear='24, Param (M)=6.0, FLOPs=485M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.652026.0496.49
MC-mini-AMC2022.0396.33
EdgeFace-SYear='24, Param (M)=3.6, FLOPs=306M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Retrained from scratch, Latency (ms) on iPhone 15 Pro=9.892026.0496.13
CircleLossTraining Dataset=clean MS1M [7], [26]2021.0996.02
CircleLoss2022.0396.02
EdgeFace-SYear='24, Param (M)=3.6, FLOPs=306M, Training Dataset=WebFace12M, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=9.892026.0495.81
PocketNetM256Year='22, Param (M)=1.75, FLOPs=1.1G, Training Dataset=CASIA-WF, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.982026.0495.66
AFRN2022.0395.56
SwiftFaceFormer-XSYear='24, Param (M)=3.4, FLOPs=294M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.592026.0495.35
SteerFaceVenue=(Ours), Generator=DM, Scale (#IDs × #Imgs)=1.2M (24K × 50)2026.0595.27
FaceLiVTv2-XSYear=-, Param (M)=2.9, FLOPs=90M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.43(35.8%↓)2026.0495.23
UIFace*Venue=ICLR 25, Generator=DM, Scale (#IDs × #Imgs)=1.2M (10K × 50)2026.0595.13
SST2022.0395.1
FaceLiVTv1-SYear='25, Param (M)=5.89, FLOPs=237M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.472026.0495.1
PocketNetM128Year='22, Param (M)=1.68, FLOPs=1.1G, Training Dataset=CASIA-WF, Model Scale=Large-Scale, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.982026.0495.07
IDPerturb*Venue=CVPR 26, Generator=DM, Scale (#IDs × #Imgs)=1.2M (24K × 50)2026.0594.92
PPFR-FDVenue=AAAI '22, PPFR=Yes2024.0394.85
MorphFaceVenue=CVPR 25, Generator=DM, Scale (#IDs × #Imgs)=1.2M (24K × 50)2026.0594.77
SteerFaceVenue=(Ours), Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0594.67
Arc2FaceVenue=ECCV 24, Generator=DM, Scale (#IDs × #Imgs)=1.2M (20K × 50, 40K × 5)2026.0594.58
EdgeFace-XSYear='24, Param (M)=1.77, FLOPs=154M, Training Dataset=WebFace12M, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=5.822026.0494.37
GhostFaceNetV2-2Year='23, Param (M)=6.8, FLOPs=77M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.672026.0494.29
UIFaceVenue=ICLR 25, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0594.29
Dyn-arcFaceTraining Dataset=clean MS1M [7], [4]2021.0994.25
PocketNetS128Year='22, Param (M)=0.92, FLOPs=587M, Training Dataset=CASIA-WF, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.882026.0494.21
MorphFaceVenue=CVPR 25, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0594.11
ProFaceVenue=MM '22, PPFR=Yes2024.0393.77
IDPerturb*Venue=CVPR 26, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0593.5
PocketNetS256Year='22, Param (M)=0.99, FLOPs=587M, Training Dataset=CASIA-WF, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.882026.0493.34
GhostFaceNetV1-2Year='23, Param (M)=4.1, FLOPs=60M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.602026.0493.31
ShuffleFaceNet-0.5Year='21, Param (M)=1.4, FLOPs=67M, Training Dataset=MS1MV2, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.452026.0492.6
AdvFaceVenue=CVPR '23, PPFR=Yes2024.0392.21
Arc2FaceVenue=ECCV 24, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0591.87
CemiFaceVenue=NeurIPS 24, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0591.06
DigiFaceVenue=WACV 23, Generator=Rendering, Scale (#IDs × #Imgs)=1.2M (10K × 72, 100K × 5)2026.0589.81
DCFaceVenue=CVPR 23, Generator=DM, Scale (#IDs × #Imgs)=1.2M (20K × 50, 40K × 5)2026.0588.61
CloakVenue=WWW '21, PPFR=Yes, Trade-off parameter=1002024.0387.97
DigiFaceVenue=WACV 23, Generator=Rendering, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0587.4
ID3Venue=NeurIPS 24, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0586.84
IDiff-FaceVenue=ICCV 23, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0585.47
DCFaceVenue=CVPR 23, Generator=DM, Scale (#IDs × #Imgs)=0.5M (10K × 50)2026.0585.33
InstaHideVenue=ICML '20, PPFR=Yes2024.0383.2