Generative Models for Pose Transfer
This addresses the problem of video synthesis for pose transfer, but it appears incremental as it builds on existing methods like pix2pix.
The paper tackles pose transfer between persons in videos, comparing nearest neighbor and generative models (pix2pix), with the generative model outperforming k-NN in generating corresponding frames and generalizing beyond demonstrated actions.
We investigate nearest neighbor and generative models for transferring pose between persons. We take in a video of one person performing a sequence of actions and attempt to generate a video of another person performing the same actions. Our generative model (pix2pix) outperforms k-NN at both generating corresponding frames and generalizing outside the demonstrated action set. Our most salient contribution is determining a pipeline (pose detection, face detection, k-NN based pairing) that is effective at perform-ing the desired task. We also detail several iterative improvements and failure modes.