Beyond Structural Unification: A Perspective and Roadmap on Unified Multimodal Models

作者：Luo Yinyi, Haoyue Bai, Yiqiao Jin, Wenyue Hua, Srijan Kumar, M. Savvides, Qingsong Wen, Jindong Wang · 发表于：Zenodo (CERN European Organization for Nuclear Research) · 年份：2026 · DOI：10.5281/zenodo.20990998 · 研究领域：Speech and dialogue systems、Language, Metaphor, and Cognition、Language and cultural evolution