A Gaussian Mixture Model (GMM)-based approach for detecting and classifying minority classes in imbalanced datasets.

A meta-learning framework for handling class imbalance in supervised learning tasks. Leverages episodic training and adaptive sampling to boost minority class performance across diverse domains.

A flexible cross-modal knowledge distillation framework for transferring supervision between heterogeneous modalities to improve model generalization and efficiency on low-resource tasks.