Everybody Dance Now

Caroline Chan
Shiry Ginosar
Tinghui Zhou
Alexei A. Efros
UC Berkeley



This paper presents a simple method for "do as I do" motion transfer: given a source video of a person dancing we can transfer that performance to a novel (amateur) target after only a few minutes of the target subject performing standard moves. We pose this problem as a per-frame image-to-image translation with spatio-temporal smoothing. Using pose detections as an intermediate representation between source and target, we learn a mapping from pose images to a target subject's appearance. We adapt this setup for temporally coherent video generation including realistic face synthesis. Our video demo can be found here.




Paper

Everybody Dance Now

Caroline Chan, Shiry Ginosar, Tinghui Zhou, Alexei A. Efros

[hosted on arXiv]

[PDF]
[Bibtex]


Video



Media





Engadget
CNET
The Verge
The Register
PC Magazine
Motherboard (Vice)
NVIDIA
Digg
Digital Trends
The Wall Street Journal
MIT Technology Review
The New Yorker
ABC Nightline


Acknowledgements

This work was supported, in part, by NSF grant IIS-1633310 and research gifts from Adobe, eBay, and Google. This webpage template was borrowed from here.