From bb471ca7c4e98d54080e96a00d86aa5c15dc3941 Mon Sep 17 00:00:00 2001 From: xingranzh <8767450+xingranzh@users.noreply.github.com> Date: Tue, 8 Jun 2021 18:04:35 +0800 Subject: [PATCH] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 500efe8..dad22ba 100644 --- a/README.md +++ b/README.md @@ -5,7 +5,7 @@ **CVPR 2021, oral presentation**
[Xingran Zhou](http://xingranzh.github.io/), [Bo Zhang](https://www.microsoft.com/en-us/research/people/zhanbo/), [Ting Zhang](https://www.microsoft.com/en-us/research/people/tinzhan/), [Pan Zhang](https://panzhang0212.github.io/), [Jianmin Bao](https://jianminbao.github.io/), [Dong Chen](https://www.microsoft.com/en-us/research/people/doch/), [Zhongfei Zhang](https://www.cs.binghamton.edu/~zhongfei/), [Fang Wen](https://www.microsoft.com/en-us/research/people/fangwen/)
Paper: https://arxiv.org/pdf/2012.02047.pdf
-Video: https://youtu.be/aBr1lOjm_FA
+ Slides: https://github.com/xingranzh/CocosNet-v2/blob/master/slides/cocosnet_v2_slides.pdf
Abstract: *We present the full-resolution correspondence learning for cross-domain images, which aids image translation. We adopt a hierarchical strategy that uses the correspondence from coarse level to guide the fine levels. At each hierarchy, the correspondence can be efficiently computed via PatchMatch that iteratively leverages the matchings from the neighborhood. Within each PatchMatch iteration, the ConvGRU module is employed to refine the current correspondence considering not only the matchings of larger context but also the historic estimates. The proposed CoCosNet v2, a GRU-assisted PatchMatch approach, is fully differentiable and highly efficient. When jointly trained with image translation, full-resolution semantic correspondence can be established in an unsupervised manner, which in turn facilitates the exemplar-based image translation. Experiments on diverse translation tasks show that CoCosNet v2 performs considerably better than state-of-the-art literature on producing high-resolution images.*