Dual-Domain Image Synthesis using Segmentation-Guided GAN

@article{Bazazian2022DualDomainIS,
  title={Dual-Domain Image Synthesis using Segmentation-Guided GAN},
  author={Dena Bazazian and Andrew Calway and Dima Damen},
  journal={2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
  year={2022},
  pages={506-515}
}
  • Dena BazazianA. CalwayD. Damen
  • Published 19 April 2022
  • Computer Science
  • 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)
We introduce a segmentation-guided approach to synthesise images that integrate features from two distinct domains. Images synthesised by our dual-domain model belong to one domain within the semantic-mask, and to another in the rest of the image - smoothly integrated. We build on the successes of few-shot StyleGAN and single-shot semantic segmentation to minimise the amount of training required in utilising two domains.The method combines few-shot cross-domain StyleGAN with a latent optimiser… 

References

SHOWING 1-10 OF 37 REFERENCES

SEAN: Image Synthesis With Semantic Region-Adaptive Normalization

We propose semantic region-adaptive normalization (SEAN), a simple but effective building block for Generative Adversarial Networks conditioned on segmentation masks that describe the semantic

Perceptual Losses for Real-Time Style Transfer and Super-Resolution

This work considers image transformation problems, and proposes the use of perceptual loss functions for training feed-forward networks for image transformation tasks, and shows results on image style transfer, where aFeed-forward network is trained to solve the optimization problem proposed by Gatys et al. in real-time.

Unsupervised Image-to-Image Translation Networks

This work makes a shared-latent space assumption and proposes an unsupervised image-to-image translation framework based on Coupled GANs that achieves state-of-the-art performance on benchmark datasets.

Illumination-Aware Multi-Task GANs for Foreground Segmentation

A robust model that allows accurately extracting the foreground even in exceptionally dark or bright scenes and in continuously varying illumination in a video sequence is presented by a triple multi-task generative adversarial network (TMT-GAN) that effectively models the semantic relationship between the dark and bright images and performs binary segmentation end-to-end.

Deep Image Blending

Image composition is an important operation to create visual content. Among image composition tasks, image blending aims to seamlessly blend an object from a source image onto a target image with

GP-GAN: Towards Realistic High-Resolution Image Blending

This paper proposes a framework called Gaussian-Poisson Generative Adversarial Network (GP-GAN), which is the first work that explores the capability of GANs in high-resolution image blending task and achieves the state-of-the-art performance on Transient Attributes dataset.

Editing in Style: Uncovering the Local Semantics of GANs

A simple and effective method for making local, semantically-aware edits to a target output image via a novel manipulation of style vectors that relies on the emergent disentanglement of semantic objects learned by StyleGAN during its training.

Repurposing GANs for One-shot Semantic Part Segmentation

This work proposes a simple and effective approach based on GANs for semantic part segmentation that requires as few as one label example along with an unlabeled dataset and produces surprisingly good results that are comparable to those from supervised baselines trained with significantly more labels.

Foreground-aware Semantic Representations for Image Harmonization

This work proposes a novel architecture to utilize the space of high-level features learned by a pre-trained classification network to harmonize composites and sets up a new state-of-the-art in terms of MSE and PSNR metrics.

Few-Shot Adaptation of Generative Adversarial Networks

This paper proposes a simple and effective method, Few-Shot GAN (FSGAN), for adapting GANs in few-shot settings (less than 100 images), which repurposes component analysis techniques and learns to adapt the singular values of the pre-trained weights while freezing the corresponding singular vectors.