Video-Text As Game Players: Hierarchical Banzhaf Interaction for Cross-Modal Representation Learning | Read Paper on Bytez