Semantic Graph Convolutional Networks for 3D Human Pose Regression

Abstract

In this paper, we study the problem of learning Graph Convolutional Networks (GCNs) for regression. Current architectures of GCNs are limited to the small receptive field of convolution filters and shared transformation matrix for each node. To address these limitations, we propose Semantic Graph Convolutional Networks (SemGCN), a novel neural network architecture that operates on regression tasks with graph-structured data. SemGCN learns to capture semantic information such as local and global node relationships, which is not explicitly represented in the graph. These semantic relationships can be learned through end-to-end training from the ground truth without additional supervision or hand-crafted rules. We further investigate applying SemGCN to 3D human pose regression. Our formulation is intuitive and sufficient since both 2D and 3D human poses can be represented as a structured graph encoding the relationships between joints in the skeleton of a human body. We carry out comprehensive studies to validate our method. The results prove that SemGCN outperforms state of the art while using 90% fewer parameters.

In CVPR 2019 (13 pages including supplementary material). The code can be found at https://github.com/garyzhao/SemGCN

References in corpus (2)

• A New Representation of Skeleton Sequences for 3D Action Recognition
• Cartoonish sketch-based face editing in videos using identity deformation transfer

Cited by in corpus (22)

• Skeleton-Based Action Recognition with Multi-Stream Adaptive Graph Convolutional Networks
• Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future
• Cascaded deep monocular 3D human pose estimation with evolutionary training data
• Vision-Based Human Pose Estimation via Deep Learning: A Survey
• Adaptive Multi-view and Temporal Fusing Transformer for 3D Human Pose Estimation
• Structured Landmark Detection via Topology-Adapting Deep Graph Learning
• Improving Robustness and Accuracy via Relative Information Encoding in 3D Human Pose Estimation
• Regular Splitting Graph Network for 3D Human Pose Estimation
• TRiPOD: Human Trajectory and Pose Dynamics Forecasting in the Wild
• Learning Dynamical Human-Joint Affinity for 3D Pose Estimation in Videos
• Graph Neural Networks in Computer Vision -- Architectures, Datasets and Common Approaches
• View-Invariant, Occlusion-Robust Probabilistic Embedding for Human Pose
• Lifting 2D Human Pose to 3D with Domain Adapted 3D Body Concept
• Dual-stream Spatio-Temporal GCN-Transformer Network for 3D Human Pose Estimation
• Learning Pyramid-structured Long-range Dependencies for 3D Human Pose Estimation
• Structure from Articulated Motion: Accurate and Stable Monocular 3D Reconstruction without Training Data
• Beyond Over-smoothing: Uncovering the Trainability Challenges in Deep Graph Neural Networks
• SkeletonX: Data-Efficient Skeleton-based Action Recognition via Cross-sample Feature Aggregation
• Monocular Human Shape and Pose with Dense Mesh-borne Local Image Features
• ActiveNet: A computer-vision based approach to determine lethargy
• MeronymNet: A Hierarchical Approach for Unified and Controllable Multi-Category Object Generation
• Alleviating Human-level Shift : A Robust Domain Adaptation Method for Multi-person Pose Estimation