dlwbm123/smoothinglatentspace

CVPR 2021

Smoothing the Disentangled Latent Style Space for Unsupervised Image-to-image Translation

[Paper] | [Poster] | [Codes]
Yahui Liu^1,3, Enver Sangineto¹, Yajing Chen², Linchao Bao², Haoxian Zhang², Nicu Sebe¹, Bruno Lepri³, Wei Wang¹, Marco De Nadai^3
1University of Trento, Italy, ²Tencent AI Lab, China, ³Bruno Kessler Foundation, Italy.
To appear in CVPR 2021
The repository offers the official implementation of our paper in PyTorch.

Image-to-Image (I2I) multi-domain translation models are usually evaluated also using the quality of their semantic interpolation results. However, state-of-the-art models frequently show abrupt changes in the image appearance during interpolation, and usually perform poorly in interpolations across domains. In this paper, we propose a new training protocol based on three specific losses which help a translation network to learn a smooth and disentangled latent style space in which: 1) Both intra- and inter-domain interpolations correspond to gradual changes in the generated images and 2) The content of the source image is better preserved during the translation. Moreover, we propose a novel evaluation metric to properly measure the smoothness of latent style space of I2I translation models. The proposed method can be plugged in existing translation approaches, and our extensive experiments on different datasets show that it can significantly boost the quality of the generated images and the graduality of the interpolations.

Our method generates smooth interpolations within and across domains in various image-to-image translation tasks.

Teaser video

Click the figure to watch the teaser video.

1.Configuration

See the environment.yml. We provide an user-friendly configuring method via Conda system, and you can create a new Conda environment using the command:

conda env create -f environment.yml

Codes will be released soon ...

2.Testing

For fast testing, we provide pretrained models on CelebA-HQ (gender) and AFHQ (animal faces):

CelebA-HQ	AFHQ
GoogleDrive	GoogleDrive

The models can be tested directly by using the offical codes of StarGAN v2.

3.Training

• Data Preparing

• Training

Acknowledgments

This code is based on the StarGAN v2. Thanks to the contributors of this project.

Citation

@inproceedings{liu2021smoothing,
  title={Smoothing the Disentangled Latent Style Space for Unsupervised Image-to-image Translation},
  author={Liu, Yahui and Sangineto, Enver and Chen, Yajing and Bao, Linchao and Zhang, Haoxian and Sebe, Nicu and Lepri, Bruno and Wang, Wei and De Nadai, Marco},
  booktitle={CVPR},
  year={2021}
}

If you have any questions, please contact me without hesitation (yahui.liu AT unitn.it).