Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

300W

Benchmarks

Task NameDataset NameSOTA ResultTrend
Facial landmark detection300W challenging
NME3.99
159
Face Alignment300W (Challenging)
NME0.0519
93
Face Alignment300W Common
NME2.52
90
Face Alignment300W Fullset (test)
NME1.42
82
Facial Landmark Detection300W
NME2.87
52
Landmark Localization300W (Chall.)
Mean Error (%)4.52
44
Landmark Localization300W (Common)
NME2.41
44
Face Alignment300W common subset
NME2.67
33
Facial landmark detection300W official (test)
AUC (0.08)58.85
32
Face Alignment300W Challenging Subset (test)
NME4.52
25
Face Alignment300W Common Subset (test)
NME2.53
25
Facial Landmark Detection300W Fullset 2013 (test)
NME2.99
21
Landmark Localization300W (full)
Mean Error (%)3.3
21
Face Alignment300W public Common (test)
Mean Error0.0473
21
Facial Landmark Localization300W (Common)
NME2.33
20
Face Alignment300W Full
NMSE3.91
19
Face Alignment300W Challenge
NMSE6.71
19
Facial Landmark Detection300W inter-pupil distance normalized (Full set)
NME4.02
19
Facial Landmark Detection300W inter-pupil distance normalized (Common set)
NME3.27
19
Face Alignment300W
NME Full2.93
18
Landmark Detection300W (Challenge)
NME4.48
17
Landmark Detection300W (Full)
NME2.96
17
Face Alignment300W public Challenging (test)
NME (pupils)7
16
Face Alignment300W Common (test)
NME (Pupils)3.67
16
Face Alignment300W Challenge (test)
NME (Inter-ocular)4.5
15
Showing 25 of 53 rows