Multimodal sentiment analysis faces significant challenges in real-world applications due to the frequent absence of modalities caused by privacy concerns, device limitations, or security policies. This article introduces a random modality dropout based on generative approach (RMDG), designed to enhance the robustness and performance of multimodal models under various modality absence scenarios. The RMDG method employs a generative approach during the training phase, where random modality dropout is applied to simulate missing modalities. By leveraging the remaining modalities to predict and regenerate the key features of the missing ones, the model effectively adapts to dynamic and unpredictable modality absences. This strategy not only eliminates the need for separate training or adjustments for each modality combination but also significantly improves the efficiency and accuracy of sentiment analysis in incomplete multimodal data scenarios. Extensive experiments demonstrate that RMDG outperforms existing methods, achieving superior performance in both complete and missing modality conditions.

While deep neural networks (DNNs)-based personalized federated learning (PFL) is demanding for addressing data heterogeneity and shows promising performance, existing methods for federated learning (FL) suffer from efficient systematic uncertainty quantification. The Bayesian DNNs-based PFL is usually questioned of either oversimplified model structures or high computational and memory costs. In this article, we introduce FedSI, a novel Bayesian DNNs-based subnetwork inference (SI) PFL framework. FedSI is simple and scalable by leveraging Bayesian methods to incorporate systematic uncertainties effectively. It implements a client-specific SI mechanism, selects network parameters with large variance to be inferred through posterior distributions, and fixes the rest as deterministic ones. FedSI achieves fast and scalable inference while preserving the systematic uncertainties to the fullest extent. Extensive experiments on four different benchmark datasets demonstrate that FedSI outperforms existing Bayesian and non-Bayesian FL baselines in heterogeneous FL scenarios.

Federated learning (FL) enables collaborative model training across distributed clients without sharing local data, providing an effective paradigm for privacy-preserving learning. To equip FL with principled uncertainty quantification, Bayesian neural networks have been introduced into federated settings. However, existing Bayesian FL frameworks typically rely on dense posterior parameterization or stochastic sampling methods, which lead to high computational and communication overhead and limit their scalability. In this paper, we propose FedKT, a nature-inspired Bayesian federated learning framework that introduces the k-tied parameterization into mean-field variational inference to achieve compact and efficient posterior approximation. By mimicking probabilistic reasoning mechanisms observed in nature, FedKT achieves adaptive and self-regularizing uncertainty modeling across resource-constrained consumer devices. Specifically, by factorizing the variational standard deviations into shared low-rank components, FedKT reduces the parameter complexity from O(mn) to O(k(m+n)), significantly improving efficiency while maintaining accurate uncertainty estimation. Extensive experiments on four benchmark datasets demonstrate that FedKT achieves comparable or superior predictive accuracy and uncertainty calibration to existing Bayesian FL methods, while ensuring low-energy, privacy-preserving, and scalable training suitable for secure consumer electronic applications.