DeepCap: Monocular Human Performance Capture Using Weak Supervision

TL;DR

DeepCap introduces a weakly supervised deep learning method for monocular dense human performance capture, eliminating the need for 3D ground truth data and outperforming existing methods in quality and robustness.

Contribution

It presents a novel weakly supervised deep learning framework that disentangles pose and surface deformation for monocular human performance capture.

Findings

Outperforms state-of-the-art in quality and robustness.

Removes need for 3D ground truth annotations.

Effective in dense, space-time coherent geometry reconstruction.

Abstract

Human performance capture is a highly important computer vision problem with many applications in movie production and virtual/augmented reality. Many previous performance capture approaches either required expensive multi-view setups or did not recover dense space-time coherent geometry with frame-to-frame correspondences. We propose a novel deep learning approach for monocular dense human performance capture. Our method is trained in a weakly supervised manner based on multi-view supervision completely removing the need for training data with 3D ground truth annotations. The network architecture is based on two separate networks that disentangle the task into a pose estimation and a non-rigid surface deformation step. Extensive qualitative and quantitative evaluations show that our approach outperforms the state of the art in terms of quality and robustness.