Augmented Feature Representation with Parallel Convolution for Cross-domain Facial Expression Recognition

Keyphrases

• Feature Representation 100%
• Generalization Performance 100%
• Multi-scale Features 100%
• Facial Expression Recognition 100%
• Augmented Features 100%
• Inconsistency 50%
• Self-attention Mechanism 50%
• Feature Distribution 50%
• Art Performance 50%
• Learning Model 50%
• Unseen 50%
• Different Datasets 50%
• FER2013 50%
• Recognition Performance 50%
• Representational Capacity 50%
• Complex Correlation 50%
• JAFFE 50%
• Cross-domain Recognition 50%
• Multiple Combinations 50%
• Cross-dataset Facial Expression Recognition 50%
• Parallel Convolution 50%
• RAF-DB 50%
• Orthogonal Regularization 50%

Computer Science

• Facial Expression 100%
• Expression Recognition 100%
• Generalization Performance 66%
• Regularization 33%
• Self-Attention Mechanism 33%
• Art Performance 33%
• Average Accuracy 33%
• Recognition Performance 33%
• Aligning Feature 33%

Engineering

• Facial Expression 100%
• Multiscale 66%
• Scale Feature 66%
• Tasks 33%
• Past Decade 33%
• Regularization 33%