Convolutional transformer network for fine-grained action recognition

Keyphrases

• 3D Convolution 20%
• Action Recognition Dataset 20%
• Convolution Operation 20%
• Convolutional 20%
• Convolutional Neural Network 40%
• Convolutional Transformer 100%
• Depthwise Separable Convolution 20%
• Diving 20%
• Dynamic Attention 20%
• Dynamic Learning 20%
• Epic 20%
• Fine-grained Recognition 100%
• Kitchen 20%
• Linear Map 20%
• Long Sequence 20%
• Long-range Dependence 20%
• Low-level Features 20%
• Multi-head Attention 20%
• Natural Language Processing 20%
• Recognition Task 20%
• Self-attention Layer 20%
• Shared Weights 20%
• Similar Action 20%
• Spatial-temporal Features 40%
• Subtle Interaction 20%
• Token 80%
• Token Embedding 20%
• Training Data 20%
• Transformer 100%
• Transformer Network 100%
• Video Clips 20%
• Video Processing 20%

Computer Science

• Action Recognition 100%
• Convolution Operation 25%
• Convolutional Neural Network 50%
• Experimental Result 25%
• Linear Mapping 25%
• Natural Language Processing 25%
• Range Dependency 25%
• Temporal Feature 50%
• Training Data 25%
• Video Processing 25%