Federated Continual Multimodal Learning (FedCMM), a framework that embeds continual-learning safeguards into the federated optimization loop at three complementary levels, is proposed, confirming that holistic, modality-aware optimization enables robust evolutive adaptation across heterogeneous networked AI deployments.
Abstract
Federated fine-tuning of Multimodal Large Language Models (MLLMs) across distributed networks enables privacy-sensitive adaptation to evolving data streams, yet a fundamental obstacle prevents robust deployment in dynamic environments: catastrophic forgetting, wherein sequential task updates erase previously acquired knowledge across visual, linguistic, and cross-modal representations. Addressing this challenge is especially critical for autonomous networked AI operating in safety-sensitive domains, such as content moderation, where reliable retention of prior knowledge underpins system integrity. To overcome this, we propose Federated Continual Multimodal Learning (FedCMM), a framework that embeds continual-learning safeguards into the federated optimization loop at three complementary levels. At the parameter level, modality-aware elastic weight consolidation computes separate Fisher information matrices for the vision encoder, language backbone, and cross-modal projector, providing granular, asymmetry-aware protection against modality-specific forgetting. At the data level, each client trains a lightweight local generative replay module to synthesize raw-data-free embedding-level multimodal replay tuples without any raw data sharing. At the aggregation level, Task-similarity-aware gradient aggregation autonomously filters and reweights client updates by gradient cosine similarity, suppressing conflicting directions and stabilizing the global learning trajectory. Extensive experiments on two benchmarks demonstrate that FedCMM consistently outperforms recent baselines on accuracy and backward transfer, confirming that holistic, modality-aware optimization enables robust evolutive adaptation across heterogeneous networked AI deployments.
This work proposes RAVEL-FCL, a generative replay-based framework for federated continual learning that integrates an improved generative model based on Rebooting ACGAN with multi-level feature alignment to ensure consistency and employs Elastic Variational Continual Learning on the server to probabilistically regularize the global model and preserve past knowledge.
Yurui Zhou, Jia Hu, G. Min et al.· ACM Transactions on Autonomo...· 0 citations
FedTaste is proposed, a parameter-efficient framework for topology-aware structural transfer in Multimodal Federated Learning with missing modalities that avoids explicit modality imputation while preserving shared semantic structure across clients.
Haocheng Liang, Jie Zhang, H. Ochiai· arXiv.org· 0 citations
Federated Graph Learning (FGL) enables privacy-preserving GNN training over distributed graph data, yet dynamic task streams in Federated Graph Continual Learning (FGCL) inevitably lead to catastrophic forgetting. From a spectral perspective, this forgetting manifests as two fundamental challenges: high-frequency inconsistency forgetting, where newly emerging node inconsistencies disrupt message passing and erase discriminative knowledge, and low-frequency consistency forgetting, where over-adaptation to new tasks dilutes global semantic coherence. Existing FGCL methods fail to explicitly address these dual spectral issues, resulting in severe degradation of knowledge retention and task generalization across continual learning stages. To this end, we propose FedFST, a spectral-aware framework that mitigates dual spectral forgetting. FedFST comprises Historical High-Frequency Knowledge Restoration (HHKR) to reconstruct and preserve high-frequency inconsistency knowledge, and Historical Low-Frequency Semantic Transfer (HLST) to stabilize low-frequency consistency via spectral distillation. Extensive experiments demonstrate the effectiveness of FedFST in alleviating catastrophic forgetting in FGCL. The code is available at https://github.com/YunQi572/FedFST.git.
Hanyao Guo, Zihan Tan, Wenke Huang et al.· Proceedings of the 32nd ACM...· 0 citations
Federated graph learning enables collaborative training over decentralized graph data without sharing raw graph information. As such risks evolve, clients must learn emerging classes from private multimodal graph streams, retain historical categories, and reject samples outside the known class space. In this setting, clients must learn emerging classes from private multimodal graph streams while preserving historical categories and rejecting samples outside the current known class space. The core challenge is catastrophic forgetting, which in federated multimodal graphs is not merely a classifier-level failure: old knowledge can be erased through modality-semantic overwriting, topology-induced structural erosion, and federated memory fragmentation. To address this challenge, we propose \textbf{FedOGL}, a semantic-structural memory preservation framework. On the client side, FedOGL preserves historical decision behavior through replay and task-start distillation, while protecting graph-propagation memory via projection onto a globally shared structure basis. On the server side, FedOGL maintains and transfers compact category prototypes to facilitate cross-client knowledge sharing without exposing raw graph data. Extensive experiments demonstrate that, compared with the best-performing baselines, FedOGL reduces performance degradation caused by catastrophic forgetting by \textbf{42.67\%}, while maintaining or improving performance on downstream tasks.
Zekai Chen, Haodong Lu, Shihao Li et al.· arXiv.org· 0 citations
FedADB, a Class Anchor-Driven Dual-Branch FL framework, a dual-branch collaborative training mechanism designed for clients that achieves significant improvements in both accuracy and convergence speed.
Zhenyan Liu, Hua Zhang, Haoran Gao et al.· 0 citations
