Enhance the Robustness of Text-Centric Multimodal Alignments

Open in new window