Client-level federated unlearning seeks to update an already trained global model so that the influence of a specified client is weakened, while the model remains effective for the remaining clients. Existing methods are mostly designed for homogeneous model settings and often rely on retraining, historical updates, or direct gradient reversal, making them less suitable for clients with heterogeneous computational capacities. This paper studies client-level unlearning in heterogeneous multi-exit federated learning, where clients participate with different model depths. We propose FedDRU, a depth-aware residual unlearning method that uses retained-client directions to represent shared knowledge, decomposes the target-client update into shared and residual components, and reverses only the residual contribution with a lightweight confidence softening constraint. Experiments on CIFAR-10 and CIFAR-100 with three-exit ResNet and ViT models show that FedDRU effectively suppresses target-client influence while maintaining competitive retained-client accuracy, achieving a stable trade-off between forgetting effectiveness and retained-client utility.
Jing-Yi Leng, Zheng-Yi Zhong, Hai-Lu Xin et al.· 2026 12th International Conf...· 0 citations
In modern operational environments, rapid and hands-free target localization is crucial for situational awareness. However, traditional plotting systems rely on cumbersome manual interactions, and conventional multimodal algorithms degrade significantly under extreme background noise and constrained communication links. To address these challenges, we propose a novel edge-cloud collaborative Speech-to-Plot (STP) framework. The proposed system integrates a domain-adapted Automatic Speech Recognition (ASR) module—fine-tuned via a noise-injected curriculum—with a zero-shot visual grounding model to translate natural voice commands into precise spatial bounding boxes. Evaluations on a custom domain-specific dataset demonstrate that our framework exhibits graceful degradation rather than severe degradation under extreme acoustic interference, maintaining robust target semantic extraction even at 0 dB Signal-to-Noise Ratio. This reliable acoustic front-end helps prevent cascading errors in downstream cross-modal attention mechanisms, enabling accurate visual target localization. Furthermore, stress testing under simulated narrowband communication networks validates the practical engineering viability of our decoupled architecture. By offloading heavy multimodal inference to the cloud, the system mitigates computational congestion, supporting operational resilience despite the inevitable physical bandwidth limitations of field deployments.
Zhong-Hao Zhou, Hai-Lu Xin, Ping Tang et al.· 2026 12th International Conf...· 0 citations
Reliable data collection is essential for disaster-oriented Internet of Things (IoT) systems, where damaged terrestrial communication infrastructure often leaves sensed data buffered at disconnected end devices. In Unmanned Aerial Vehicle (UAV)-Internet of Things device (IoTD) collaborative data collection, random UAV faults and limited energy and buffer resources further complicate mission execution, making fault-tolerant scheduling crucial for robust data recovery. To address these issues, a unified framework is developed by integrating dynamic UAV reliability modeling, Maximum Distance Separable (MDS)-coded fault-tolerant backup, and collaborative scheduling optimization. Within this framework, a data fault-tolerance mechanism, termed MFTB, and a bilevel collaborative scheduling algorithm, termed LP-DCFS, are proposed. Simulation results indicate that, in the evaluated scenarios, the proposed methods achieve better overall performance than the considered baselines. In a representative high-load, high-failure scenario, MFTB reduces data loss by 4.8% and 37.5% compared with Buffer-Limited Retransmission (BLR) and Replication, respectively, while LP-DCFS increases the amount of recovered data by 33.9%, 32.3%, and 53.1% compared with ACEPSO, ADE-DMRM, and DQN, respectively. Under the modeled independent random crash and non-return faults and the evaluated simulation settings, these results suggest that coordinating failure-risk characterization, data-protection mechanisms, and task-scheduling strategies can improve the robustness and data-recovery capability of disaster-oriented UAV-assisted data collection.
Hailu Xin, Weidong Bao, Hui Yan et al.· Drones· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
