Robust facial landmark detection by cross-order cross-semantic deep network

Computer Science

• Landmark Detection 100%
• Semantic Feature 71%
• Detection Method 42%
• Convolutional Neural Network 28%
• Representation Learning 28%
• Experimental Result 14%
• Pose Variation 14%

Keyphrases

• Cross-order 100%
• Facial Landmark Detection 100%
• Multi-excitation 18%
• Semantic Feature Learning 9%
• Channel Correlation 9%
• Discriminative Representation Learning 9%
• Multiple Attention 9%