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

3FabRec: Fast Few-shot Face alignment by Reconstruction

About

Current supervised methods for facial landmark detection require a large amount of training data and may suffer from overfitting to specific datasets due to the massive number of parameters. We introduce a semi-supervised method in which the crucial idea is to first generate implicit face knowledge from the large amounts of unlabeled images of faces available today. In a first, completely unsupervised stage, we train an adversarial autoencoder to reconstruct faces via a low-dimensional face embedding. In a second, supervised stage, we interleave the decoder with transfer layers to retask the generation of color images to the prediction of landmark heatmaps. Our framework (3FabRec) achieves state-of-the-art performance on several common benchmarks and, most importantly, is able to maintain impressive accuracy on extremely small training sets down to as few as 10 images. As the interleaved layers only add a low amount of parameters to the decoder, inference runs at several hundred FPS on a GPU.

Bjoern Browatzki, Christian Wallraven• 2019

Related benchmarks

TaskDatasetResultRank
Face AlignmentWFLW (test)
NME (%) (Testset)5.62
144
Face Alignment300W (Challenging)
NME0.0574
93
Face Alignment300W Common
NME3.36
90
Face Alignment300-W (Full)
NME3.82
66
Landmark Localization300W (Chall.)
Mean Error (%)5.74
44
Landmark Localization300W Common
NME3.36
44
Face AlignmentAFLW Frontal
NME (%)1.59
22
Facial Landmark DetectionCHARLOTTE High 1.0
NME (W/H)0.0814
18
Landmark Localization300W Private 1.0 (test)
FR10% Error0.0017
12
Face AlignmentAFLW Full
NME (%)1.84
11
Showing 10 of 15 rows

Other info

Code

Follow for update