arXiv：cs.AI· Liangjian Wen, Linjie Li, Jiang Duan, Yong Dai, Jianzhuang Liu, Zhao Kang·· 6 小时前
依赖、压缩与协同：多模态学习的统一信息论视角
Dependency, Compression, and Synergy: A Unified Information-Theoretic View of Multimodal Learning
AI 导读
一篇综述从信息论视角统一了多模态学习，将互信息、信息瓶颈与部分信息分解分别对应模态间依赖、任务导向的信息压缩保留以及冗余/独有/协同的信息拆解。该综述回顾了 2018–2026 年的 170 项近期研究与 12 项基础工作，围绕跨模态对齐、信息高效融合、交互类型刻画和扩展至多模态基础模型四大挑战组织内容。
正文
Abstract:Recent advances in multimodal foundation models have intensified the need to understand how different modalities share, preserve, and complement information. Mutual Information (MI), the Information Bottleneck (IB), and Partial Information Decomposition (PID) provide complementary perspectives, yet existing studies often treat them as isolated tools. This survey presents an information-theoretic perspective connecting these principles as progressively refined views of multimodal information processing: MI characterizes inter-modal dependency, IB explains task-oriented information preservation under compression, and PID decomposes preserved information into redundancy, uniqueness, and synergy. We review 170 recent studies (2018--2026) and 12 foundational works, organizing multimodal learning around four challenges: cross-modal alignment, information-efficient fusion, interaction-type characterization, and scaling to multimodal foundation models. Rather than using application domains as primary taxonomy axes, we interpret healthcare, robotics, recommendation systems, affective computing, and wireless communications as empirical validations of these information principles. Beyond taxonomy, we organize existing multimodal paradigms within a single information-theoretic coordinate system -- the Generalized Multimodal Information Lagrangian -- in which they occupy exact or approximate parameter corners, and whose unoccupied regions name candidate method families the literature has not yet built. We further discuss how emerging multimodal foundation models instantiate these principles at scale and identify open challenges including scalable information estimation in high-dimensional settings, standardized evaluation across information-theoretic methods, combinatorial complexity of multimodal PID, and the transition from post-hoc information analysis toward information-aware multimodal learning.
| Subjects: | Information Theory (cs.IT); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.14421 [cs.IT] |
| (or arXiv:2609.14421v1 [cs.IT] for this version) | |
| https://doi.org/10.48550/arXiv.2609.14421 arXiv-issued DOI via DataCite |
Submission history
From: Liangjian Wen PhD. [view email]
[v1]
Sun, 13 Sep 2026 10:43:11 UTC (573 KB)
来源：arXiv：cs.AI · arxiv.org