Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Past month
  • Any time
  • Past hour
  • Past 24 hours
  • Past week
  • Past month
  • Past year
All results
18 hours ago · We propose a spatial-structural graph convolution method which can better represent symmetrical human structures and edge nodes. ... Rethinking the st-gcns for 3d ...
Missing: behavior | Show results with:behavior
Jul 3, 2024 · ST-GCN is the first GCN algorithm working on 3D skeleton ... Skeleton-based human action recognition with global context-aware attention LSTM networks.
5 days ago · In practice, we develop a novel pipeline that extracts skeleton coordinates using pose estimation and tracking algorithms and employ Multi-person Panoramic GCN ...
Missing: behavior | Show results with:behavior
Jul 19, 2024 · This method has proven effective in improving human action recognition. Several variants derived from ST-GCN have achieved remarkable results [4], [5], [6], ...
Missing: behavior | Show results with:behavior
Jul 11, 2024 · This demonstrates the beneficial impact of depth estimation and dimensional enhancement. Substituting the CNN with a GCN for depth estimation in 2D-SCHAR (D-GM) ...
Missing: behavior | Show results with:behavior
Jul 17, 2024 · Core algorithm ST-GCN consists of BN (batch normalization), GCN (graph convolutional network), TCN (temporal convolutional network), POOL (pooling), and FC ( ...
Jul 22, 2024 · The proposed model outperforms the STGCN model, achieving an 11.19% increase in MAE, a 12.37% improvement in RMSE and a 4.97% reduction in inference time. These ...
6 days ago · In the distracted driving behaviour recognition based on body posture, this section designs an ablation experiment using ST-GCN as the baseline, and unfolds the ...
1 day ago · The study employs a hybrid network model of CNN and BLSTM to obtain. IB features that contain group relationships through graph convolutional networks (GCN) and ...
Jul 12, 2024 · Following the overall goals of our framework, we want to create one framework that unifies GNNs across the spatial and spectral domains as well as within each ...