Data Augmentation for Human Motion Prediction
Published in 2021 17th International Conference on Machine Vision Applications (MVA), 2021
Recommended citation: T. Maeda and N. Ukita, "Data Augmentation for Human Motion Prediction," 2021 17th International Conference on Machine Vision and Applications (MVA), 2021, pp. 1-5, doi: 10.23919/MVA51890.2021.9511368.
Human motion prediction is seldom deployed to real-world tasks due to difficulty in collecting a huge amount of motion data. We propose two motion data augmentation approaches using Variational AutoEncoder (VAE) and Inverse Kinematics (IK). Our VAE-based generative model with adversarial training and sampling near samples generates various motions even with insufficient original motion data. Our IK-based augmentation scheme allows us to semi-automatically generate a variety of motions. Furthermore, we correct unrealistic artifacts in the augmented motions. As a result, our method outperforms previous motion augmentation methods.
Results