Self-Supervised Multimodal Learning

Cross-modal Reconstruction

Learning objective aimed at reconstructing a complete modality from a partial representation or another modality.

← Tillbaka