Semi-Supervised Few-Shot Learning from A Dependency-Discriminant Perspective

Keyphrases

• Unlabeled Data 100%
• Few-shot Learning 100%
• Semi-supervised Few-shot Learning 100%
• Support Set 50%
• Learning Approaches 50%
• ImageNet 50%
• Pseudo-labeling 50%
• Few-shot Task 50%
• Learning Problems 25%
• Mini 25%
• Learning Methods 25%
• Learning Performance 25%
• Unseen 25%
• Discriminant Analysis 25%
• Statistical Dependence 25%
• Loss-based 25%
• Training Tasks 25%
• Classification Benchmark 25%
• Embedding Features 25%
• Label Prediction 25%
• Cross-covariance Operator 25%
• Hilbert-Schmidt Norm 25%
• Meta-learning Paradigm 25%
• Dependence Maximization 25%
• Few-shot Classification 25%
• Inductive Bias 25%
• Labeled Training Data 25%

Computer Science

• Few-Shot Learning 100%
• Unlabeled Data 66%
• Learning Approach 33%
• Learning Problem 16%
• Discriminant Analysis 16%
• Training Data 16%
• Labeled Example 16%
• Learning Performance 16%
• Meta-Learning 16%