Human Motion Modeling using DVGANs
This work addresses motion modeling for applications like animation or robotics, but it appears incremental as it builds on existing GAN methods.
The paper tackles human motion modeling by proposing a novel generative adversarial network (GAN) with a dense validation discriminator and translation invariance, achieving results in motion generation and completion as evaluated on Human 3.6M and CMU datasets using inception scores.
We present a novel generative model for human motion modeling using Generative Adversarial Networks (GANs). We formulate the GAN discriminator using dense validation at each time-scale and perturb the discriminator input to make it translation invariant. Our model is capable of motion generation and completion. We show through our evaluations the resiliency to noise, generalization over actions, and generation of long diverse sequences. We evaluate our approach on Human 3.6M and CMU motion capture datasets using inception scores.