Skip to content

You Said That?: Synthesising Talking Faces from Audio

Notifications You must be signed in to change notification settings

ifgcguitarclub/yousaidthat

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Prerequisites

ffmpeg
cuda (v8.0)

For face alignment:

scikit-image
dlib (v18.18)
numpy
scipy

Setting up Matconvnet

cd matconvnet_gen
addpath matlab
vl_compilenn('enableGpu', true, ...
               'cudaRoot', '/usr/local/cuda-8.0', ...
               'cudaMethod', 'nvcc') ;
vl_testnn('gpu', true)

Demo

Download model:

sh download_model.sh

Run demo in MATLAB:

run_demo

Face alignment

cd face_detection
y = baseface('../data/obama.jpg');

5 aligned identity images are then concatenated channel-wise. See faceimg.mat for examples.

Citation

Please cite the paper below if you make use of the demo.

@InProceedings{Chung17b,
  author       = "Chung, J.~S. and Jamaludin, A. and Zisserman, A.",
  title        = "You said that?",
  booktitle    = "British Machine Vision Conference",
  year         = "2017",
}

About

You Said That?: Synthesising Talking Faces from Audio

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • MATLAB 35.6%
  • Cuda 30.9%
  • C++ 13.7%
  • TeX 9.1%
  • Python 5.9%
  • Shell 1.9%
  • Other 2.9%