Learning Global Pose Features in Graph Convolutional Networks for 3D Human Pose Estimation

Liu, Kenkun; Zou, Zhiming; Tang, Wei

doi:10.1007/978-3-030-69525-5_6

Cited by 17 publications

(11 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although the model brings strong constraints, the iterative optimization used to solve the dictionary weights or SMPL parameters is time-consuming. GCN (Cai et al, 2019;Liu et al, 2020;Zhao et al, 2019) and Attention (Guo et al, 2021) are used to capture the relationship between two joints. In (Yang et al, 2022;Chen et al, 2019b;Habibie et al, 2019), the encoder-decoder is applied to create the latent space which is used to mine the inter-dependencies between joints.…”

Section: Related Workmentioning

confidence: 99%

View Consistency Aware Holistic Triangulation for 3D Human Pose Estimation

Wan¹,

Chen²,

Zhao³

2023

Preprint

View full text Add to dashboard Cite

The rapid development of multi-view 3D human pose estimation (HPE) is attributed to the maturation of monocular 2D HPE and the geometry of 3D reconstruction. However, 2D detection outliers in occluded views due to neglect of view consistency, and 3D implausible poses due to lack of pose coherence, remain challenges. To solve this, we introduce a Multi-View Fusion module to refine 2D results by establishing view correlations. Then, Holistic Triangulation is proposed to infer the whole pose as an entirety, and anatomy prior is injected to maintain the pose coherence and improve the plausibility. Anatomy prior is extracted by PCA whose input is skeletal structure features, which can factor out global context and joint-by-joint relationship from abstract to concrete. Benefiting from the closed-form solution, the whole framework is trained end-to-end. Our method outperforms the state of the art in both precision and plausibility which is assessed by a new metric.

show abstract

Section: Related Workmentioning

confidence: 99%

View Consistency Aware Holistic Triangulation for 3D Human Pose Estimation

Wan¹,

Chen²,

Zhao³

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Fig. 1 shows the prediction accuracy of recent 3D pose estimation methods [24,32,46,20,51]. As can be observed from this figure, most of them suffer from poor performance for some complex actions, such as, "Sit," "Sit Down," and "Take Photos."…”

Section: Observation and Motivationmentioning

confidence: 99%

“…Please noted that the HCSF layer could also adapt to other graph-based frameworks, like GCN [11], Pre-Agg, Post-Agg [20], the difference lies in weight sharing schemes, which is orthogonal to our approach.…”

Section: Hierarchical Channel-squeezing Fusion Layermentioning

confidence: 99%

“…To tackle this problem, geometric dependencies are incorporated into the network in [7,28,39,46], which significantly improve prediction accuracy. As articulated human body can be naturally modeled as a graph, with the recent development of graph neural networks (GNN) [11,13,1,41,47], various GNN-based methods [49,6,20,2,51] are proposed in the literature for 2D-to-3D pose estimation.…”

Section: Introductionmentioning

confidence: 99%

“…For example, for poses in "running", the hand-foot joints are closely related, while for poses in "Sitting", there is no such strong relationship. It is rather difficult to capture such information with a static FCN [24] Multi-task [23] SRNet [46] Global-GCN [20] High-order GCN [51] (5) Complex Pose (6) Rare Pose (4) Self-occlusion (3) Depth Ambiguity (1) Easy Poses…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation

Zeng¹,

Sun²,

Yang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Various deep learning techniques have been proposedto solve the single-view 2D-to-3D pose estimation problem. While the average prediction accuracy has been improved significantly over the years, the performance on hard poses with depth ambiguity, self-occlusion, and complex or rare poses is still far from satisfactory. In this work, we target these hard poses and present a novel skeletal GNN learning solution. To be specific, we propose a hop-aware hierarchical channel-squeezing fusion layer to effectively extract relevant information from neighboring nodes while suppressing undesired noises in GNN learning.In addition, we propose a temporal-aware dynamic graph construction procedure that is robust and effective for 3D pose estimation. Experimental results on the Human3.6M dataset show that our solution achieves 10.3% average prediction accuracy improvement and greatly improves on hard poses over state-of-the-art techniques. We further apply the proposed technique on the skeleton-based action recognition task and also achieve state-of-the-art performance. Our code is available at https://github. com/ailingzengzzz/Skeletal-GNN .

show abstract