Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks

By Harsh Rai, Naman Shukla, Ziyu Zhou. 2018

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Generative Models

Citations in the corpus

CNN-Generated Images Are Surprisingly Easy to Spot… for Now (2019, methodology): “[50], each trained from one of the two image sources (CycleGAN and AutoGAN), and one of the two image representations (images and spectrum) respectively.”
Generative Adversarial Networks (2021, positive): “Meanwhile, the style transfer utilized high level semantic features, which highly improve the quality of the output image and save a great amount of time for the producer [9].”
Learning Layout and Style Reconfigurable GANs for Controllable Image Synthesis (2020, background): “, class information [3], [5], [6], [9], [36], source image [10], [12], [13], [37], text description [16], [17], [38], [39], scene graph [18], [40], etc) as input.”
Semantic Image Synthesis With Spatially-Adaptive Normalization (2019, background): “The GANs consist of a generator and a discriminator where the goal of the generator is to produce realistic images so that the discriminator cannot tell the synthesized images apart from real ones.”
High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs (2017, methodology): “We adopt the same evaluation protocol from previous image-to-image translation works [21, 63].”
StarGAN: Unified Generative Adversarial Networks for Multi-domain Image-to-Image Translation (2017, methodology): “We leverage PatchGANs [7, 14, 32] for the discriminator network, which classifies whether local image patches are real or fake.”
A Variational Inequality Perspective on Generative Adversarial Networks (2018, background): “Generative adversarial networks (GANs) (Goodfellow et al., 2014) form a generative modeling approach known for producing realistic natural images (Karras et al., 2018) as well as high quality super-resolution (Ledig et al., 2017) and style transfer (Zhu et al., 2017).”
Generative Feature Replay For Class-Incremental Learning (2020, background): “Generative adversarial networks (GANs) [11] are able to generate realistic and sharp images conditioned on object categories [12, 39], text [42, 57], another image (image translation) [18, 58] and style transfer [10].”
A Papier-Mache Approach to Learning 3D Surface Generation (2018, neutral): “Significant progress has been made on learning good representations for images, allowing impressive applications in image generation [16, 34].”
End-to-End Recovery of Human Shape and Pose (2017, background): “Since these meshes do not necessarily have a corresponding image, we refer to this data as unpaired [55].”
The Perception-Distortion Tradeoff (2017, background): “GAN-based image restoration Most recently, GAN-based methods have demonstrated unprecedented perceptual quality in super-resolution [3], [6], [9], [37], inpainting [7], [20], [38], compression [21], [39], [40], deblurring [41] and image-to-image translation [11], [22], [42].”
Moment Matching for Multi-Source Domain Adaptation (2018, background): “moment matching [47], adversarial domain confusion [40, 8, 38] and GAN-based alignment [50, 15, 23].”
The Effectiveness of Data Augmentation in Image Classification using Deep Learning (2017, methodology): “Next, we will make use of CycleGAN [19] to augment our data by transferring styles from images in the dataset to a fixed predetermined image such as Night/Day theme or Winter/Summer.”
CyCADA: Cycle-Consistent Adversarial Domain Adaptation (2017, methodology): “More recently, Cycle-consistent Adversarial Networks (CycleGAN) (Zhu et al., 2017) produced compelling image translation results such as generating photorealistic images from impressionism paintings or transforming horses into zebras at high resolution using the cycle-consistency loss.”
StarGAN v2: Diverse Image Synthesis for Multiple Domains (2019, methodology): “Early image-to-image translation methods [20, 53, 29] are well known to learn a deterministic mapping even with stochastic noise inputs.”