Object-Region Video Transformers

Keyphrases

• Action Recognition 16%
• Attention Module 16%
• Capture Trajectory 16%
• Dynamic Modules 16%
• Epic 16%
• Few-shot Action Recognition 16%
• Object Dynamics 33%
• Object Information 16%
• Object Region 100%
• Object Representation 50%
• Object-centric 16%
• Object-level 16%
• Performance Improvement 16%
• Region Attention 16%
• Self-attention 16%
• Spatio-temporal Representation 16%
• Temporal Action Detection 16%
• Token 16%
• Transformer Architecture 16%
• Transformer Block 16%
• Transformer Layer 33%
• Transformer Model 16%
• Two-stream 16%
• Video Understanding 16%
• Vision Transformer 100%
• Visual Object 16%

Computer Science

• Action Recognition 100%
• Convolutional Neural Network 50%
• Information Object 50%
• Object Representation 100%
• Performance Improvement 50%
• Video Understanding 50%