A deep generative model that describes human motions can benefit a wide range of fundamental computer vision and graphics tasks, such as providing robustness to video-based human pose estimation, predicting complete body movements for motion capture systems during occlusions, and assisting key frame animation with plausible movements. In this paper, we present a method for learning complex human motions independent of specific tasks using a combined global and local latent space to facilitate coarse and fine-grained modeling. Specifically, we propose a hierarchical motion variational autoencoder (HM-VAE) that consists of a 2-level hierarchical latent space. While the global latent space captures the overall global body motion, the local latent space enables to capture the refined poses of the different body parts. We demonstrate the effectiveness of our hierarchical motion variational autoencoder in a variety of tasks including video-based human pose estimation, motion completion from partial observations, and motion synthesis from sparse key-frames. Even though, our model has not been trained for any of these tasks specifically, it provides superior performance than task-specific alternatives. Our general-purpose human motion prior model can fix corrupted human body animations and generate complete movements from incomplete observations.
Our general purpose motion prior consists of a latent space of human motions and is learned using a hierarchical motion variational autoencoder (HM-VAE). Our approach is task-generic and can be directly adopted to a wide range of applications. Left: Motion interpolation and completion can be accomplished by traversing the latent space. Right: Noisy pose estimation can be refined by projecting noisy inputs into our latent space and decoding back. And a latent vector in the learned latent space is corresponding to a valid motion sequence.