# TraVeLGAN: Image-to-image Translation by Transformation Vector Learning Source: [TraVeLGAN: Image-to-image Translation by Transformation Vector Learning]() Collection index: [llms.txt]() Data updated: 2025-10-25T21:25:53.160Z Authors: Matthew Amodio, Smita Krishnaswamy Published: 2019-02-25T21:48:32.000Z ## Abstract Interest in image-to-image translation has grown substantially in recent years with the success of unsupervised models based on the cycle-consistency assumption. The achievements of these models have been limited to a particular subset of domains where this assumption yields good results, namely homogeneous domains that are characterized by style or texture differences. We tackle the challenging problem of image-to-image translation where the domains are defined by high-level shapes and contexts, as well as including significant clutter and heterogeneity. For this purpose, we introduce a novel GAN based on preserving intra-domain vector transformations in a latent space learned by a siamese network. The traditional GAN system introduced a discriminator network to guide the generator into generating images in the target domain. To this two-network system we add a third: a siamese network that guides the generator so that each original image shares semantics with its generated version. With this new three-network system, we no longer need to constrain the generators with the ubiquitous cycle-consistency restraint. As a result, the generators can learn mappings between more complex domains that differ from each other by large differences - not just style or texture. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image-to-Image Translation on AFHQ Dog -\> Cat (test)](): Image-to-Image Translation; AFHQ Dog -\> Cat (test); FID: 58.91. - [Image-to-Image Translation on CelebA-HQ & AFHQ Human Face -\> Cat (test)](): Image-to-Image Translation; CelebA-HQ & AFHQ Human Face -\> Cat (test); FID: 85.28. - [Image-to-Image Translation on ImageNet-291 (Bird -\> Dog) (test)](): Image-to-Image Translation; ImageNet-291 (Bird -\> Dog) (test); FID: 169.98. - [Image-to-Image Translation on ImageNet-291 Bird -\> Car (test)](): Image-to-Image Translation; ImageNet-291 Bird -\> Car (test); FID: 164.28. - [Image-to-Image Translation on User Study](): Image-to-Image Translation; User Study; Content Consistency: 0.012. - [Image-to-Image Translation on CelebA-HQ Male \<-\> Female (test)](): Image-to-Image Translation; CelebA-HQ Male \<-\> Female (test); FID: 66.6. ## Original sources and implementation [Original paper]() [PDF]()