# ManiFest: Manifold Deformation for Few-shot Image Translation

Source: [ManiFest: Manifold Deformation for Few-shot Image Translation](<https://www.sota2.com/research/paper/68fa6d181e209131cea37ac3>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T00:25:36.384Z

Authors: Fabio Pizzati, Jean-Fran\\c{c}ois Lalonde, Raoul de Charette

Published: 2021-11-26T18:59:58.000Z

## Abstract

Most image-to-image translation methods require a large number of training images, which restricts their applicability. We instead propose ManiFest: a framework for few-shot image translation that learns a context-aware representation of a target domain from a few images only. To enforce feature consistency, our framework learns a style manifold between source and proxy anchor domains (assumed to be composed of large numbers of images). The learned manifold is interpolated and deformed towards the few-shot target domain via patch-based adversarial and feature statistics alignment losses. All of these components are trained simultaneously during a single end-to-end loop. In addition to the general few-shot translation task, our approach can alternatively be conditioned on a single exemplar image to reproduce its specific style. Extensive experiments demonstrate the efficacy of ManiFest on multiple tasks, outperforming the state-of-the-art on all metrics and in both the general- and exemplar-based scenarios. Our code is available at https://github.com/cv-rits/Manifest .

## 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 Editing on our dataset film-grey style](<https://www.sota2.com/research/sota/image-editing-on-our-dataset-film-grey-style.md>): Image Editing; our dataset film-grey style; PSNR: 17.83.

- [Image Editing on film-dream-blue style](<https://www.sota2.com/research/sota/image-editing-on-film-dream-blue-style.md>): Image Editing; film-dream-blue style; PSNR: 14.63.

- [Style Editing on Style Editing Dataset isp style](<https://www.sota2.com/research/sota/style-editing-on-style-editing-dataset-isp-style.md>): Style Editing; Style Editing Dataset isp style; PSNR: 14.9.

- [Style Editing on StyleQoRA lomo style (test)](<https://www.sota2.com/research/sota/style-editing-on-styleqora-lomo-style-test.md>): Style Editing; StyleQoRA lomo style (test); PSNR: 13.94.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2111.13681>)

[PDF](<https://arxiv.org/pdf/2111.13681.pdf>)
