Towards Robust RGB-D Human Mesh Recovery
arXiv:1911.07383
Abstract
We consider the problem of human pose estimation. While much recent work has focused on the RGB domain, these techniques are inherently under-constrained since there can be many 3D configurations that explain the same 2D projection. To this end, we propose a new method that uses RGB-D data to estimate a parametric human mesh model. Our key innovations include (a) the design of a new dynamic data fusion module that facilitates learning with a combination of RGB-only and RGB-D datasets, (b) a new constraint generator module that provides SMPL supervisory signals when explicit SMPL annotations are not available, and (c) the design of a new depth ranking learning objective, all of which enable principled model training with RGB-D data. We conduct extensive experiments on a variety of RGB-D datasets to demonstrate efficacy.
arXiv comment10 pages, 4 figures, 4 tables
References in corpus (17)
- Deep Residual Learning for Image Recognition
- Learning from Synthetic Humans
- Stacked Hourglass Networks for Human Pose Estimation
- MoCap-guided Data Augmentation for 3D Pose Estimation in the Wild
- PKU-MMD: A Large Scale Benchmark for Continuous Multi-Modal Human Action Understanding
- A simple yet effective baseline for 3d human pose estimation
- Synthesizing Training Images for Boosting Human 3D Pose Estimation
- End-to-end Recovery of Human Shape and Pose
- Weakly-Supervised Discovery of Geometry-Aware Representation for 3D Human Pose Estimation
- Unite the People: Closing the Loop Between 3D and 2D Human Representations
- Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation
- Integral Human Pose Regression
- Learning elementary structures for 3D shape generation and matching
- Shape-Aware Human Pose and Shape Reconstruction Using Multi-View Images
- Vision-based Pose Estimation for Augmented Reality : A Comparison Study
- 3D-CODED : 3D Correspondences by Deep Deformation
- A Multi-view RGB-D Approach for Human Pose Estimation in Operating Rooms