Няма описание

chenxuanhong ad5029eba5 update преди 4 години
data 01a8d6d0a6 init преди 4 години
doc ad5029eba5 update преди 4 години
models 01a8d6d0a6 init преди 4 години
options 01a8d6d0a6 init преди 4 години
util 01a8d6d0a6 init преди 4 години
.gitattributes a0dab0c7b1 Initial commit преди 4 години
.gitignore 01a8d6d0a6 init преди 4 години
LICENSE a0dab0c7b1 Initial commit преди 4 години
README.md ad5029eba5 update преди 4 години
test_one_image.py 01a8d6d0a6 init преди 4 години
train.py 01a8d6d0a6 init преди 4 години

README.md

SimSwap: An Efficient Framework For High Fidelity Face Swapping

Proceedings of the 28th ACM International Conference on Multimedia

The official repository with Pytorch

Currently only the test code is available, training scripts are coming soon

[Conference paper]

Results

Results1

Results2

Video

High-quality videos can be found in the link below:

[Baidu Drive link for video] Password: b26n

Dependencies

  • python3.6+
  • pytorch1.5+
  • torchvision
  • opencv
  • pillow
  • numpy

Usage

Use python3.5, pytorch1.3.0

Use this command to test the face swapping between two images:

python test_one_image.py --isTrain false --name people --Arc_path models/BEST_checkpoint.tar --pic_a_path crop_224/mars.jpg --pic_b_path crop_224/ds.jpg --output_path output/

--name refers to the checkpoint name.

Pretrained model

[Google Drive] [Baidu Drive]

To cite our paper

@inproceedings{DBLP:conf/mm/ChenCNG20,
  author    = {Renwang Chen and
               Xuanhong Chen and
               Bingbing Ni and
               Yanhao Ge},
  title     = {SimSwap: An Efficient Framework For High Fidelity Face Swapping},
  booktitle = {{MM} '20: The 28th {ACM} International Conference on Multimedia},
  pages     = {2003--2011},
  publisher = {{ACM}},
  year      = {2020},
  url       = {https://doi.org/10.1145/3394171.3413630},
  doi       = {10.1145/3394171.3413630},
  timestamp = {Thu, 15 Oct 2020 16:32:08 +0200},
  biburl    = {https://dblp.org/rec/conf/mm/ChenCNG20.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}