Neural Kinematic Networks for Unsupervised Motion Retargetting

TL;DR

This paper introduces a neural network architecture with a forward kinematics layer and cycle consistency training for unsupervised motion retargeting, enabling adaptation to different characters without paired data.

Contribution

It presents a novel recurrent neural network with a forward kinematics layer and cycle consistency loss for unsupervised, online motion retargeting across diverse characters.

Findings

Achieves state-of-the-art results on Mixamo data

Works online, adapting motion sequences in real-time

Successfully retargets motion from monocular videos to 3D characters

Abstract

We propose a recurrent neural network architecture with a Forward Kinematics layer and cycle consistency based adversarial training objective for unsupervised motion retargetting. Our network captures the high-level properties of an input motion by the forward kinematics layer, and adapts them to a target character with different skeleton bone lengths (e.g., shorter, longer arms etc.). Collecting paired motion training sequences from different characters is expensive. Instead, our network utilizes cycle consistency to learn to solve the Inverse Kinematics problem in an unsupervised manner. Our method works online, i.e., it adapts the motion sequence on-the-fly as new frames are received. In our experiments, we use the Mixamo animation data to test our method for a variety of motions and characters and achieve state-of-the-art results. We also demonstrate motion retargetting from monocular human videos to 3D characters using an off-the-shelf 3D pose estimator.