List your company
ResearchFormerly WizWand

Facial Landmark Detection on AFLW Full

0.0156NMEAnchorFace
-0.2077761.3000122.80784.315588Mar 12, 2018Dec 7, 2018Sep 4, 2019Jun 1, 2020Feb 27, 2021Nov 25, 2021Aug 23, 2022
Updated 8mo ago

Evaluation results

MethodLinks
AnchorFacebackbone=ShuffleNet-V2, resolution=224x2242020.070.0156
HGs + SA + GHCUSemantic Alignment=true, Global Heatmap Correction Unit=true2019.030.016
HGs + SASemantic Alignment=true2019.030.0162
SA2020.070.0162
ODN2020.070.0163
Wing2019.030.0165
Wing2020.070.0165
DSRN2019.030.0186
SAN2018.030.0191
SAN2020.070.0191
HGs2019.030.0195
SBR2019.030.0214
Two-Stage2018.030.0217
TSR2019.030.0217
DCFE2019.030.0217
TSR2020.070.0217
CCL2018.030.0272
CCL2019.030.0272
CCL2020.070.0272
CFSS2018.030.0392
CFSS2019.030.0392
CFSS2020.070.0392
SDM2018.030.0405
LBF2020.070.0424
LBF2018.030.0425
LBF2019.030.0425
ERT2018.030.0435
LABbackbone=Hourglass, extra_info=B2019.041.25
LAB (w/ B)backbone=Hourglass, extra boundary information=true2019.081.25
LAB (w/ B)Backbone=Hourglass, extra_info=boundary information2022.081.25
LABYear=2018, Backbone=ResNet-18, External area data for training=true2020.031.25
WBF → HRNetV2Backbone=HRNetV2-W18, pre-training=WBF dataset2022.081.42
PIPNetBackbone=ResNet-1012020.031.42
PIPNetBackbone=ResNet-502020.031.44
PDBbackbone=ResNet-50, extra_info=DA2019.041.47
PDB (w/ DA)backbone=ResNet-50, stronger data augmentation=true2019.081.47
PDB (w/ DA)Backbone=ResNet-50, extra_info=stronger data augmentation2022.081.47
WingYear=2018, Backbone=ResNet-502020.031.47
PIPNetBackbone=ResNet-182020.031.48
PIPNetBackbone=MobileNetV22020.031.52
PIPNetBackbone=MobileNetV32020.031.52
HRNetV2-W18backbone=HRNetV2-W182019.041.57
HRNetV2-W18backbone=HRNetV2-W182019.081.57
HRNetV2Backbone=HRNetV2-W182022.081.57
HRNetYear=2019, Backbone=HRNetV2-W182020.031.57
RCN+Year=2018, External area data for training=true2020.031.59
HG+SA+GHCUYear=2019, Backbone=Hourglass2020.031.6
ODNYear=2019, Backbone=ResNet-182020.031.63
LABbackbone=Hourglass, extra_info=none2019.041.85
LAB (w/o B)backbone=Hourglass, extra boundary information=false2019.081.85
LAB (w/o B)Backbone=Hourglass2022.081.85
LABYear=2018, Backbone=ResNet-182020.031.85
DSRN2019.041.86
DSRN2019.081.86
DSRNBackbone=-2022.081.86
SANbackbone=ResNet-1522019.041.91
SANbackbone=ResNet-1522019.081.91
SANBackbone=ResNet-1522022.081.91
SANYear=2018, Backbone=ITN-CPM2020.031.91
LaplaceKLYear=20192020.031.97
3DDEYear=2019, External area data for training=true2020.032.01
CPM + SBRbackbone=CPM2019.042.14
CPM + SBRbackbone=CPM2019.082.14
CPM + SBRBackbone=CPM2022.082.14
TSRbackbone=VGG-S2019.042.17
DCFEextra_info=3D2019.042.17
TSRbackbone=VGG-S2019.082.17
DCFE (w/ 3D)extra 3D information=true2019.082.17
TSRBackbone=VGG-S2022.082.17
DCFE (w/ 3D)Backbone=-, extra_info=3D information2022.082.17
TSRYear=20172020.032.17
DCFEYear=2018, External area data for training=true2020.032.17
DAC-CSR2019.042.27
DAC-CSR2019.082.27
DAC-CSRBackbone=-2022.082.27
DAC-CSRYear=20172020.032.27
CCL2019.042.72
CCL2019.082.72
CCLBackbone=-2022.082.72
RCPR2019.043.73
RCPR2019.083.73
RCPRBackbone=-2022.083.73
CFSS2019.043.92
CFSS2019.083.92
CFSSBackbone=-2022.083.92
SDM2019.044.05
SDM2019.084.05
SDMBackbone=-2022.084.05
LBF2019.044.25
LBF2019.084.25
LBFBackbone=-2022.084.25
ERT2019.044.35
ERT2019.084.35
ERTBackbone=-2022.084.35
CDM2019.045.43
CDM2019.085.43
CDMBackbone=-2022.085.43
RCN2019.045.6
RCN2019.085.6
RCNBackbone=-2022.085.6