CVPR 2020 oral Cross-domain Correspondence Learning for Exemplar-based Image Translation
Pan Zhang Bo Zhang Dong Chen Lu Yuan Fang Wen
University of Science and Technology of China Microsoft Research Asia


We propose an exemplar-based image synthesis. Given the exemplar images (1st row), our network translates the inputs in the form of segmentation mask, edge and pose, to photorealistic images (2nd row) under the guidance of dense correspondence to the exemplar image.



We present a general framework for exemplar-based image translation, which synthesizes a photo-realistic image from the input in a distinct domain (e.g., semantic segmentation mask, or edge map, or pose keypoints), given an exemplar image. The output has the style (e.g., color, texture) in consistency with the semantically corresponding objects in the exemplar. We propose to jointly learn the cross domain correspondence and the image translation, where both tasks facilitate each other and thus can be learned with weak supervision. The images from distinct domains are first aligned to an intermediate domain where dense correspondence is established. Then, the network synthesizes images based on the appearance of semantically corresponding patches in the exemplar. We demonstrate the effectiveness of our approach in several image translation tasks. Our method is superior to state-of-the-art methods in terms of image quality significantly, with the image style faithful to the exemplar with semantic consistency. Moreover, we show the utility of our method for several applications.


Network architecture

The illustration of the CoCosNet architecture. Given the input $x_A\in A$ and the exemplar $y_B \in B$, the correspondence submodule adapts them into the same domain $S$, where dense correspondence can be established. Then, the translation network generates the final output based on the warped exemplar $r_{y\to x}$ according to the correspondence, yielding an exemplar-based translation output.




We can use different exemplars to synthesis different outputs which have the style (e.g., color, texture) in consistency with the semantically corresponding objects in exemplars.


Makeup Transfer

Given a portrait and makeup strokes (1st column), we can transfer these makeup edits to other portraits by matching the semantic correspondence. We show more examples in the supplementary material.


Image Editing

Given the input image and its mask, we can semantically edit the image content through the manipulation on the mask.




"Cross-domain Correspondence Learning for Exemplar-based Image Translation",
Pan Zhang, Bo Zhang, Dong Chen, Lu Yuan and Fang Wen
Conference on Computer Vision and Pattern Recongnition (CVPR), 2020, Oral Presentation

[PDF] [Code] [Supplementary Doc.][Slides] [BibTeX]


Last updated: April 2020