# Weakly-supervised 3D Human Pose Estimation with Cross-view U-shaped Graph Convolutional Network

arXiv: https://arxiv.org/abs/2105.10882 · HTML: https://mlsanity.com/paper/2105.10882/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 40 references that are in the map.
- First posted on arXiv 23 May 2021.

## Builds on (most cited first, 29 of 29)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Stacked Hourglass Networks for Human Pose Estimation](https://mlsanity.com/paper/1603.06937/) (2016)
- [3D human pose estimation in video with temporal convolutions and semi-supervised training](https://mlsanity.com/paper/1811.11742/) (2018)
- [A simple yet effective baseline for 3d human pose estimation](https://mlsanity.com/paper/1705.03098/) (2017)
- [Cascaded Pyramid Network for Multi-Person Pose Estimation](https://mlsanity.com/paper/1711.07319/) (2017)
- [Coarse-to-Fine Volumetric Prediction for Single-Image 3D Human Pose](https://mlsanity.com/paper/1611.07828/) (2016)
- [Learnable Triangulation of Human Pose](https://mlsanity.com/paper/1905.05754/) (2019)
- [Epipolar Transformers](https://mlsanity.com/paper/2005.04551/) (2020)
- [Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach](https://mlsanity.com/paper/1704.02447/) (2017)
- [3D Human Pose Estimation = 2D Pose Estimation + Matching](https://mlsanity.com/paper/1612.06524/) (2016)
- [Self-Supervised Learning of 3D Human Pose using Multi-view Geometry](https://mlsanity.com/paper/1903.02330/) (2019)
- [Graph Stacked Hourglass Networks for 3D Human Pose Estimation](https://mlsanity.com/paper/2103.16385/) (2021)
- [SRNet: Improving Generalization in 3D Human Pose Estimation with a Split-and-Recombine Approach](https://mlsanity.com/paper/2007.09389/) (2020)
- [Exploiting Temporal Contexts with Strided Transformer for 3D Human Pose Estimation](https://mlsanity.com/paper/2103.14304/) (2021)
- [Cross View Fusion for 3D Human Pose Estimation](https://mlsanity.com/paper/1909.01203/) (2019)
- [Exploiting temporal information for 3D pose estimation](https://mlsanity.com/paper/1711.08585/) (2017)
- [RepNet: Weakly Supervised Training of an Adversarial Reprojection Network for 3D Human Pose Estimation](https://mlsanity.com/paper/1902.09868/) (2019)
- [Motion Guided 3D Pose Estimation from Videos](https://mlsanity.com/paper/2004.13985/) (2020)
- [Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image](https://mlsanity.com/paper/1701.00295/) (2017)
- [Weakly-Supervised 3D Human Pose Learning via Multi-view Images in the Wild](https://mlsanity.com/paper/2003.07581/) (2020)
- [Lightweight Multi-View 3D Pose Estimation through Camera-Disentangled Representation](https://mlsanity.com/paper/2004.02186/) (2020)
- [Harvesting Multiple Views for Marker-less 3D Human Pose Annotations](https://mlsanity.com/paper/1704.04793/) (2017)
- [Self-Supervised 3D Human Pose Estimation via Part Guided Novel Image Synthesis](https://mlsanity.com/paper/2004.04400/) (2020)
- [Compressed Volumetric Heatmaps for Multi-Person 3D Pose Estimation](https://mlsanity.com/paper/2004.00329/) (2020)
- [Articulated Pose Estimation by a Graphical Model with Image Dependent Pairwise Relations](https://mlsanity.com/paper/1407.3399/) (2014)
- [Weakly-Supervised Discovery of Geometry-Aware Representation for 3D Human Pose Estimation](https://mlsanity.com/paper/1903.08839/) (2019)
- [Rethinking Pose in 3D: Multi-stage Refinement and Recovery for Markerless Motion Capture](https://mlsanity.com/paper/1808.01525/) (2018)
- [Distill Knowledge from NRSfM for Weakly Supervised 3D Pose Learning](https://mlsanity.com/paper/1908.06377/) (2019)

## Built on by (most cited first, 13 of 13)

- [Self-supervised Action Representation Learning from Partial Spatio-Temporal Skeleton Sequences](https://mlsanity.com/paper/2302.09018/) (2023)
- [Zero-shot Skeleton-based Action Recognition via Mutual Information Estimation and Maximization](https://mlsanity.com/paper/2308.03950/) (2023)
- [GraphMLP: A Graph MLP-Like Architecture for 3D Human Pose Estimation](https://mlsanity.com/paper/2206.06420/) (2022)
- [Contrastive Learning from Spatio-Temporal Mixed Skeleton Sequences for Self-Supervised Skeleton-Based Action Recognition](https://mlsanity.com/paper/2207.03065/) (2022)
- [Regular Splitting Graph Network for 3D Human Pose Estimation](https://mlsanity.com/paper/2305.05785/) (2023)
- [Deep learning for 3D human pose estimation and mesh recovery: A survey](https://mlsanity.com/paper/2402.18844/) (2024)
- [Zero-Shot Skeleton-based Action Recognition with Dual Visual-Text Alignment](https://mlsanity.com/paper/2409.14336/) (2024)
- [LLMI3D: MLLM-based 3D Perception from a Single 2D Image](https://mlsanity.com/paper/2408.07422/) (2024)
- [STGFormer: Spatio-Temporal GraphFormer for 3D Human Pose Estimation in Video](https://mlsanity.com/paper/2407.10099/) (2024)
- [Enhancing 3D Human Pose Estimation Amidst Severe Occlusion with Dual Transformer Fusion](https://mlsanity.com/paper/2410.04574/) (2024)
- [Exploring Self-supervised Skeleton-based Action Recognition in Occluded Environments](https://mlsanity.com/paper/2309.12029/) (2023)
- [Learning Pyramid-structured Long-range Dependencies for 3D Human Pose Estimation](https://mlsanity.com/paper/2506.02853/) (2025)
- [Skeleton-Snippet Contrastive Learning with Multiscale Feature Fusion for Action Localization](https://mlsanity.com/paper/2512.16504/) (2025)

## Abstract (arXiv)

Although monocular 3D human pose estimation methods have made significant progress, it is far from being solved due to the inherent depth ambiguity. Instead, exploiting multi-view information is a practical way to achieve absolute 3D human pose estimation. In this paper, we propose a simple yet effective pipeline for weakly-supervised cross-view 3D human pose estimation. By only using two camera views, our method can achieve state-of-the-art performance in a weakly-supervised manner, requiring no 3D ground truth but only 2D annotations. Specifically, our method contains two steps: triangulation and refinement. First, given the 2D keypoints that can be obtained through any classic 2D detection methods, triangulation is performed across two views to lift the 2D keypoints into coarse 3D poses. Then, a novel cross-view U-shaped graph convolutional network (CV-UGCN), which can explore the spatial configurations and cross-view correlations, is designed to refine the coarse 3D poses. In particular, the refinement progress is achieved through weakly-supervised learning, in which geometric and structure-aware consistency checks are performed. We evaluate our method on the standard benchmark dataset, Human3.6M. The Mean Per Joint Position Error on the benchmark dataset is 27.4 mm, which outperforms existing state-of-the-art methods remarkably (27.4 mm vs 30.2 mm).