Abstract
We demonstrate the benefit of using an ultimate skip (US) connection for facial expression synthesis using generative adversarial networks (GAN). A direct connection transfers identity, facial, and color details from input to output while suppressing artifacts. The intermediate layers can therefore focus on expression generation only. This leads to a light-weight US-GAN model comprised of encoding layers, a single residual block, decoding layers, and an ultimate skip connection from input to output. US-GAN has fewer parameters than state-of-the-art models and is trained on orders of magnitude smaller dataset. It yields increase in face verification score (FVS) and decrease in average content distance (ACD). Based on a randomized user-study, US-GAN outperforms the state of the art by in face realism, in expression quality, and in identity preservation.