Transformer-based spatial-Temporal feature lifting for 3D hand mesh reconstruction

Computer Science

• Consecutive Frame 100%
• Depth Information 100%
• Experimental Result 100%
• Positional Encoding 100%
• Spatial Relationship 100%
• Temporal Feature 100%
• Temporal Information 100%
• Temporal Relationship 100%
• Video Sequences 100%

Keyphrases

• 3D Hand Reconstruction 100%
• Confidence-based 33%
• Decoder 33%
• Depth Information 33%
• Feature Representation 100%
• Positional Encoding 33%
• Softmax 33%
• Spatial Relationship 33%
• Spatial-temporal Features 100%
• Temporal Relations 33%
• Transformer Architecture 33%
• Transformer Encoder 66%
• Transformer-based 100%
• Video Sequences 33%