List your company
ResearchFormerly WizWand

Face Recognition on IJB-B (TAR @ FAR=1e-4)

97.84TAR @ FAR=1e-4ResNet200-TopoFR
2.139226.984651.8376.6754Sep 21, 2021Jun 25, 2022Mar 29, 2023Dec 31, 2023Oct 3, 2024Jul 7, 2025Apr 10, 2026
Updated 6mo ago

Evaluation results

MethodLinks
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=12.242026.0497.84
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=OOM2026.0496.73
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=18.202026.0496.46
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=14.312026.0496.12
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=3.762026.0495.3
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× ↓)2026.0495.18
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%↓)2026.0495.02
LOTR-R+Face detector=PyramidKey [74]2021.0994.54
LOTR-R+Face detector=RetinaFace [67]2021.0994.54
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%↓)2026.0494.51
LOTR-M+Face detector=PyramidKey [74]2021.0994.49
LOTR-MFace detector=RetinaFace [67]2021.0994.48
LOTR-M+Face detector=RetinaFace [67]2021.0994.48
LOTR-MFace detector=PyramidKey [74]2021.0994.47
RetinaFaceFace landmarks=From detector2021.0994.43
PyramidKeyFace landmarks=From detector2021.0994.37
ArcFaceVenue=CVPR '19, PPFR=No2024.0394.13
FaceLiVTv1-MYear='25, Param (M)=9.8, FLOPs=386M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.672026.0493.7
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=9.982026.0493.69
DuetFaceVenue=MM '22, PPFR=Yes2024.0393.66
PartialFaceVenue=ICCV '23, PPFR=Yes2024.0393.64
EdgeFace-SYear='24, Param (M)=3.6, FLOPs=306M, Training Dataset=WebFace12M, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=9.892026.0493.59
MinusFaceVenue=ours, PPFR=Yes, Backbone=IR-502024.0393.37
DCTDPVenue=ECCV '22, PPFR=Yes, Privacy Budget=epsilon=12024.0393.29
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=9.982026.0493.26
GhostFaceNetV1-1Year='23, Param (M)=4.1, FLOPs=216M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.782026.0493.12
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=9.892026.0493.03
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=6.522026.0492.95
PPFR-FDVenue=AAAI '22, PPFR=Yes2024.0392.93
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=2.592026.0492.92
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=0.832026.0492.9
MobileFaceNetYear='21, Param (M)=0.99, FLOPs=440M, Training Dataset=MS1MV2, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.772026.0492.83
EdgeFace-XSYear='24, Param (M)=1.77, FLOPs=154M, Training Dataset=WebFace12M, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=5.822026.0492.67
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=0.692026.0492.3
GhostFaceNetV2-2Year='23, Param (M)=6.8, FLOPs=77M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.672026.0491.89
SwiftFaceFormer-L1Year='24, Param (M)=11.8, FLOPs=805M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=1.202026.0491.81
SwiftFaceFormer-SYear='24, Param (M)=6.0, FLOPs=485M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.652026.0491.56
MixFaceNets-MYear='21, Param (M)=3.9, FLOPs=626M, Training Dataset=MS1MV2, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.702026.0491.55
GhostFaceNetV1-2Year='23, Param (M)=4.1, FLOPs=60M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.602026.0491.25
FaceLiVTv1-SYear='25, Param (M)=5.89, FLOPs=237M, Training Dataset=Glint360K, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.472026.0491.2
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=0.982026.0490.74
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%↓)2026.0490.67
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=0.982026.0490.63
SwiftFaceFormer-XSYear='24, Param (M)=3.4, FLOPs=294M, Training Dataset=MS1MV3, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.592026.0490.2
PocketNetS128Year='22, Param (M)=0.92, FLOPs=587M, Training Dataset=CASIA-WF, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.882026.0489.44
PocketNetS256Year='22, Param (M)=0.99, FLOPs=587M, Training Dataset=CASIA-WF, Model Scale=Lightweight, Experimental Setup=Original, Latency (ms) on iPhone 15 Pro=0.882026.0489.31
AdvFaceVenue=CVPR '23, PPFR=Yes2024.0370.21
ProFaceVenue=MM '22, PPFR=Yes2024.0369.39
InstaHideVenue=ICML '20, PPFR=Yes2024.0361.88
CloakVenue=WWW '21, PPFR=Yes, Trade-off parameter=1002024.0333.58
PEEPVenue=CS '20, PPFR=Yes, Privacy Budget=52024.035.82