动量引导的联邦分裂蒸馏用于个性化时序边缘智能
Momentum-Guided Federated Split Distillation for Personalized Temporal Edge Intelligence
浏览论文内容
中文总结 AI 辅助
本文提出动量引导的联邦分裂蒸馏框架,结合TeRR-SAtt与AMGF,在智能建筑数据上显著降低边缘训练/推理延迟与资源占用,并提升本地学习精度。
中文摘要 AI 辅助
我们提出了一种动量引导的联邦分裂蒸馏框架,用于实现个性化、高效且自主的时序边缘智能。我们引入了TeRR-SAtt,这是一种新颖的时序储层学生注意力设计,它结合了固定储层表示、轻量级时序学生以及个性化输出模块。我们还提出了AMGF,即预期动量引导融合机制,该机制通过学习动量对客户端进行聚类,并推导出专门的教师更新。在真实世界的智能建筑数据上,与所考虑的基线相比,TeRR-SAtt将边缘训练延迟降低了65.50%,推理延迟降低了44.70%,训练内存使用量降低了18.40%,推理CPU使用量降低了33.10%。同时,与全局更新相比,AMGF在RMSE上最多将本地学习提升了35.31%。
英文摘要
We propose a momentum-guided federated split distillation framework for personalized, efficient, and autonomous temporal edge intelligence. We introduce TeRR-SAtt, our novel temporal reservoir student attention design that combines fixed reservoir representations, a lightweight temporal student, and personalized output modules. We also present AMGF, our anticipatory momentum-guided fusion mechanism that clusters clients through learning momentum and derives specialized teacher updates. On real-world smart-building data, TeRR-SAtt reduces edge training latency by 65.50%, inference latency by 44.70%, training memory usage by 18.40%, and inference CPU usage by 33.10% over the considered baselines. At the same time, AMGF improves local learning by up to 35.31% in RMSE compared to global updates.
发表机构
- CESI(法国高等工业研究中心)
- CESI LINEACT UR 7527(法国高等工业研究中心LINEACT联合研究单位UR 7527)
机构由 AI 辅助整理,请以论文原文为准。