Contrastive-Collapsed Loss for Flexible and Geometrically Optimal Embeddings and Faster Convergence

Part of

Machine Learning for Multimodal Medical Prediction