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

Keyphrases

• Deep Network 100%
• Cross-order 100%
• Facial Landmark Detection 100%
• Semantic Features 36%
• Detection Method 27%
• Neural Network 18%
• Regularizer 18%
• Convolutional Neural Network 18%
• Multi-excitation 18%
• Benchmark Dataset 9%
• Local Details 9%
• Learning Attention 9%
• Facial Parts 9%
• Pose Variation 9%
• Large Pose 9%
• Semantic Feature Learning 9%
• Accurately Model 9%
• Channel Correlation 9%
• Discriminative Representation Learning 9%
• Partial Occlusion 9%
• Multiple Attention 9%

Computer Science

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