View of A Modular Framework for Robust Multimodal Representation Learning via Dynamic Modality Weighting | Journal of Computer Technology and Software