Everybody Dance Now

TL;DR

This paper introduces a simple yet effective method for transferring dance motions from a source to a target person using pose-based video translation, along with a forensics tool and an open dataset.

Contribution

It presents a novel pose-to-appearance translation approach for motion transfer, a face synthesis pipeline, and releases a new dataset for training and evaluation.

Findings

Produces compelling dance motion transfer results

Includes a forensics tool to detect synthetic videos

Provides an open-source dataset for training

Abstract

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 approach this problem as video-to-video translation using pose as an intermediate representation. To transfer the motion, we extract poses from the source subject and apply the learned pose-to-appearance mapping to generate the target subject. We predict two consecutive frames for temporally coherent video results and introduce a separate pipeline for realistic face synthesis. Although our method is quite simple, it produces surprisingly compelling results (see video). This motivates us to also provide a forensics tool for reliable synthetic content detection, which is able to distinguish videos synthesized by our system from real data. In addition, we release a first-of-its-kind open-source dataset of videos that can be legally used for training and motion transfer.