Self-MI: Efficient Multimodal Fusion via Self-Supervised Multi-Task Learning with Auxiliary Mutual Information Maximization

Self-MI: Efficient Multimodal Fusion via Self-Supervised Multi-Task Learning with Auxiliary Mutual Information Maximization Cam-Van Thi Nguyen author Ngoc-Hoa Thi Nguyen author Duc-Trong Le author Quang-Thuy Ha author 2023-12 text Proceedings of the 37th Pacific Asia Conference on Language, Information and Computation Chu-Ren Huang editor Yasunari Harada editor Jong-Bok Kim editor Si Chen editor Yu-Yin Hsu editor Emmanuele Chersoni editor Pranav A editor Winnie Huiheng Zeng editor Bo Peng editor Yuxi Li editor Junlin Li editor Association for Computational Linguistics Hong Kong, China conference publication thi-etal-2023-self https://aclanthology.org/2023.paclic-1.58/ 2023-12 582 590