Home

Identity From Here, Pose From There: Self-Supervised Disentanglement and Generation of Objects Using Unlabeled Videos The proposed model takes as input an ID image and a pose image, and generates an output image with the identity of the ID image and the pose of the pose image. Methods The generator takes as input both the identity referen...

SinGAN: Learning a Generative Model From a Single Natural Image(Best Paper Award) Project Page Code Multi-Scale Pipeline Our model consists of a pyramid of GANs, where both training and inference are done in a coarse-to-fine fashion. At each scale, Gn learns to generate image samples in which all the overlappin...

Make a Face: Towards Arbitrary High Fidelity Face Manipulation (a) Our framework (b) 2D and 3D projection of 5000 facial structure representations. Each color denotes a cluster, a more intuitive illustration of our approach is to map each cluster to a Gaussian prior. extending the capacity of C-VAE constrained by single prior. Additive Memor...

Info Title: Few-shot Video-to-Video Synthesis Task: Video-to-Video Translation Author: Ting-Chun Wang, Ming-Yu Liu, Andrew Tao, Guilin Liu, Jan Kautz, Bryan Catanzaro Date: Oct. 2019 Published: NIPS 2019 Abstract Video-to-video synthesis (vid2vid) aims at converting an input semantic video, such as videos of human poses or segment...

Info Title: Few-Shot Unsupervised Image-to-Image Translation Task: Image-to-Image Translation Author: Ming-Yu Liu Xun Huang Arun Mallya Tero Karras Timo Aila Jaakko Lehtinen Jan Kautz Date: May 2019 Arxiv: 1905.01723 Published: ICCV 2019 Abstract Unsupervised image-to-image translation methods learn to map imag...

StructureFlow: Image Inpainting via Structure-Aware Appearance Flow Structure(edge-preserved smoothing) Generator + Apperace Flow for Texture Generation Overview of our StructureFlow. Our model first generates global structures (i.e. edge-preserved smooth images) using structure reconstructor. Then texture generator is used to yield high-frequ...

Guided Image-to-Image Translation With Bi-Directional Feature Transformation In this work, the authors propose to apply the conditioning operation in both direction with information flowing not only from the guidance image to the input image, but from the input image to the guidance image as well. Second, they extend the existing feature-wise f...

Info Title: SinGAN: Learning a Generative Model From a Single Natural Image Task: Image Synthesis Author: Tamar Rott Shaham, Tali Dekel, Tomer Michaeli Date: May 2019 Arxiv: 1905.01164 Published: ICCV 2019(Best Paper Award) Highlights Zero-Shot learning pipeline from single image Wild applications for super-resolution, pain...