Facial Landmark Detection Using Generative Adversarial Network Combined with Autoencoder for Occlusion

Table 1

NRMSE(×10⁻²) comparisons of our proposed model with different modules on challenging set.

Model	NRMSE
Autoencoder + DRN	7.23
GAN-IA + DRN	6.96
GAN-IA + PFLD	6.71
GAN-IA(without SA) + DRN	7.15