Team:
Joon Son Chung
Amir Jamaludin
Andrew Zisserman
https://youtu.be/lXhkxjSJ6p8
You said that? - Synthesizing videos of talking faces from audio
Published on May 8, 2017
Quote:
"You said that?"
by Joon Son Chung, Amir Jamaludin and Andrew Zisserman
Abstract:
We present a method for generating a video of a talking face. The method takes as inputs: (i) one still image of the target face, and (ii) an audio speech segment; and outputs a video of the target face lip synched with the audio. The method works in real time, and at run time, is applicable to previously unseen faces and audio (i.e. not part of the training data).
To achieve this we propose an encoder-decoder CNN model that uses a joint embedding of the face and audio to generate synthesised talking face video frames. The model is trained on tens of hours of unlabelled videos.
We also show results of re-dubbing videos using speech from a different person.
Note:
We are releasing this video at a proof of concept stage — there is clearly much to improve (in terms of graphics and post-processing) before the method reaches production quality video.
Article "AI can doctor videos to put words in the mouths of speakers"
by Timothy Revell
May 19, 2017