no code implementations • 2 Jan 2022 • Jireh Jam, Connah Kendrick, Vincent Drouard, Kevin Walker, Moi Hoon Yap
The RSTL layer easily adapts dual encoders by increasing the unique semantic information through direct communication.
1 code implementation • 18 Oct 2021 • Ricard Durall, Jireh Jam, Dominik Strassel, Moi Hoon Yap, Janis Keuper
We then incorporate the geometry information of a segmentation mask to provide a fine-grained manipulation of facial attributes.
no code implementations • 7 May 2021 • Jireh Jam, Connah Kendrick, Vincent Drouard, Kevin Walker, Moi Hoon Yap
It introduces the use of foreground segmentation masks to preserve the fidelity.
2 code implementations • 11 Aug 2020 • Jireh Jam, Connah Kendrick, Vincent Drouard, Kevin Walker, Gee-Sern Hsu, Moi Hoon Yap
We address the problem by proposing a Wasserstein GAN combined with a new reverse mask operator, namely Reverse Masking Network (R-MNet), a perceptual adversarial network for image inpainting.
no code implementations • 11 Jan 2020 • Jireh Jam, Connah Kendrick, Vincent Drouard, Kevin Walker, Gee-Sern Hsu, Moi Hoon Yap
Additionally, we propose a Wasserstein-Perceptual loss function to preserve colour and maintain realism on a reconstructed image.