# US-GAN: On the importance of Ultimate Skip Connection for Facial Expression Synthesis

arXiv: https://arxiv.org/abs/2112.13002 · HTML: https://mlsanity.com/paper/2112.13002/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 44 references that are in the map.
- First posted on arXiv 24 Dec 2021.

## Builds on (most cited first, 23 of 23)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Densely Connected Convolutional Networks](https://mlsanity.com/paper/1608.06993/) (2016)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [Perceptual Losses for Real-Time Style Transfer and Super-Resolution](https://mlsanity.com/paper/1603.08155/) (2016)
- [Conditional Generative Adversarial Nets](https://mlsanity.com/paper/1411.1784/) (2014)
- [Instance Normalization: The Missing Ingredient for Fast Stylization](https://mlsanity.com/paper/1607.08022/) (2016)
- [StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation](https://mlsanity.com/paper/1711.09020/) (2017)
- [Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks](https://mlsanity.com/paper/1703.10593/) (2017)
- [AffectNet: A Database for Facial Expression, Valence, and Arousal Computing in the Wild](https://mlsanity.com/paper/1708.03985/) (2017)
- [DualGAN: Unsupervised Dual Learning for Image-to-Image Translation](https://mlsanity.com/paper/1704.02510/) (2017)
- [FSRNet: End-to-End Learning Face Super-Resolution with Facial Priors](https://mlsanity.com/paper/1711.10703/) (2017)
- [STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing](https://mlsanity.com/paper/1904.09709/) (2019)
- [Invertible Conditional GANs for image editing](https://mlsanity.com/paper/1611.06355/) (2016)
- [RelGAN: Multi-Domain Image-to-Image Translation via Relative Attributes](https://mlsanity.com/paper/1908.07269/) (2019)
- [Learning Spatial Attention for Face Super-Resolution](https://mlsanity.com/paper/2012.01211/) (2020)
- [ExprGAN: Facial Expression Editing with Controllable Expression Intensity](https://mlsanity.com/paper/1709.03842/) (2017)
- [Cascade EF-GAN: Progressive Facial Expression Editing with Local Focuses](https://mlsanity.com/paper/2003.05905/) (2020)
- [Learning Residual Images for Face Attribute Manipulation](https://mlsanity.com/paper/1612.05363/) (2016)
- [High-Fidelity and Arbitrary Face Editing](https://mlsanity.com/paper/2103.15814/) (2021)
- [Pixel-based Facial Expression Synthesis](https://mlsanity.com/paper/2010.14397/) (2020)
- [Masked Linear Regression for Learning Local Receptive Fields for Facial Expression Synthesis](https://mlsanity.com/paper/2011.09104/) (2020)

## Built on by (most cited first, 4 of 4)

- [A Review of Human Emotion Synthesis Based on Generative Technology](https://mlsanity.com/paper/2412.07116/) (2024)
- [Facial Expression Video Generation Based-On Spatio-temporal Convolutional GAN: FEV-GAN](https://mlsanity.com/paper/2210.11182/) (2022)
- [FacEnhance: Facial Expression Enhancing with Recurrent DDPMs](https://mlsanity.com/paper/2406.09040/) (2024)
- [Improving Generative Adversarial Network Generalization for Facial Expression Synthesis](https://mlsanity.com/paper/2603.15648/) (2026)

## Abstract (arXiv)

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 3× fewer parameters than state-of-the-art models and is trained on 2 orders of magnitude smaller dataset. It yields 7% increase in face verification score (FVS) and 27% decrease in average content distance (ACD). Based on a randomized user-study, US-GAN outperforms the state of the art by 25% in face realism, 43% in expression quality, and 58% in identity preservation.