arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.08040cs.LGcs.AR
收藏

面向微控制器的低能耗端侧个性化方法

Low-Energy On-Device Personalization for MCUs

Imperial College London(帝国理工学院);University of York(约克大学)

Yushan Huang, Ranya Aloufi, Xavier Cadet, Yuchen Zhao, Payam Barnaghi, Hamed Haddadi

AI总结:

本文提出MicroT,一种基于自监督知识蒸馏和阶段决策早退推理的低能耗MCU端侧个性化方法,在保持SRAM低于1MB的同时显著降低训练与推理能耗,并优于传统迁移学习和现有SOTA方法。

摘要:

微控制器单元(MCU)因其低成本和低能耗而成为边缘应用的理想平台,并被广泛用于各类应用,包括个性化机器学习任务,其中定制化模型可以增强任务适应性。然而,现有的本地端侧个性化方法大多仅支持简单的机器学习架构,或需要复杂的本地预训练/训练,导致高能耗,削弱了MCU的低能耗优势。本文提出MicroT,一种高效且低能耗的MCU个性化方法。MicroT包含一个鲁棒、通用但极小的特征提取器,该提取器通过自监督知识蒸馏开发而成,用于训练任务特定的头部,从而以极低的能耗和计算需求实现独立的端侧个性化。MicroT实现了一种面向MCU优化的早退推理机制,称为阶段决策(stage-decision),以进一步降低能耗成本。该机制支持用户可配置的退出标准(阶段决策比率),从而在能耗成本与模型性能之间实现自适应平衡。我们使用两个模型、三个数据集和两块MCU开发板对MicroT进行了评估。在两个模型和三个数据集上,MicroT相比传统迁移学习(TTL)和两种最先进(SOTA)方法性能提升了2.12%至11.60%。针对广泛使用的能耗敏感型边缘设备,MicroT的端侧训练无需额外的复杂操作,与SOTA方法相比,能耗成本最多降低2.28倍,同时保持SRAM占用低于1MB。在本地推理过程中,与TTL相比,MicroT在两块开发板和两个数据集上将能耗成本降低了14.17%,凸显了其在能耗敏感、资源受限的MCU上长期使用的适用性。

英文摘要:

Microcontroller Units (MCUs) are ideal platforms for edge applications due to their low cost and energy consumption, and are widely used in various applications, including personalized machine learning tasks, where customized models can enhance the task adaptation. However, existing approaches for local on-device personalization mostly support simple ML architectures or require complex local pre-training/training, leading to high energy consumption and negating the low-energy advantage of MCUs. In this paper, we introduce $MicroT$, an efficient and low-energy MCU personalization approach. $MicroT$ includes a robust, general, but tiny feature extractor, developed through self-supervised knowledge distillation, which trains a task-specific head to enable independent on-device personalization with minimal energy and computational requirements. MicroT implements an MCU-optimized early-exit inference mechanism called stage-decision to further reduce energy costs. This mechanism allows for user-configurable exit criteria (stage-decision ratio) to adaptively balance energy cost with model performance. We evaluated MicroT using two models, three datasets, and two MCU boards. $MicroT$ outperforms traditional transfer learning (TTL) and two SOTA approaches by 2.12 - 11.60% across two models and three datasets. Targeting widely used energy-aware edge devices, MicroT's on-device training requires no additional complex operations, halving the energy cost compared to SOTA approaches by up to 2.28X while keeping SRAM usage below 1MB. During local inference, MicroT reduces energy cost by 14.17% compared to TTL across two boards and two datasets, highlighting its suitability for long-term use on energy-aware resource-constrained MCUs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑