LifeLearner：面向嵌入式计算平台的硬件感知元持续学习系统
LifeLearner: Hardware-Aware Meta Continual Learning System for Embedded Computing Platforms
AI总结：
本文提出LifeLearner，一个硬件感知的元持续学习系统，通过结合元学习、重放策略及无损有损压缩技术，在确保高准确率的同时大幅降低内存、延迟和能耗，成功在资源受限的嵌入式边缘设备上实现高效持续学习。
摘要：
持续学习（CL）允许用户个性化和家庭机器人等应用在运行中学习并适应上下文。当上下文、动作和用户发生变化时，这是一项重要功能。然而，由于标注数据、内存和计算能力有限，在资源受限的嵌入式系统上启用CL具有挑战性。在本文中，我们提出了LifeLearner，这是一个硬件感知的元持续学习系统，它在确保高准确率的同时，极大地优化了系统资源（更低的内存、延迟和能耗）。具体而言，我们（1）利用元学习和重放策略来明确应对数据稀缺问题并确保高准确率；（2）有效结合无损和有损压缩，显著降低CL和重放样本的资源需求；（3）考虑硬件特性，在嵌入式和物联网平台上开发了硬件感知系统。因此，LifeLearner实现了接近最优的CL性能，与Oracle基线相比，准确率仅低2.8%。与最先进的（SOTA）Meta CL方法相比，LifeLearner大幅减少了内存占用（降低178.7倍），端到端延迟降低了80.8-94.2%，能耗降低了80.9-94.2%。此外，我们成功将LifeLearner部署在两个边缘设备和一个微控制器单元上，从而在资源受限的平台上实现了高效的CL，而在这些平台上运行SOTA方法是不切实际的，这促进了可适应CL的广泛普及。代码可在https://github.com/theyoungkwon/LifeLearner获取。
英文摘要：
Continual Learning (CL) allows applications such as user personalization and household robots to learn on the fly and adapt to context. This is an important feature when context, actions, and users change. However, enabling CL on resource-constrained embedded systems is challenging due to the limited labeled data, memory, and computing capacity. In this paper, we propose LifeLearner, a hardware-aware meta continual learning system that drastically optimizes system resources (lower memory, latency, energy consumption) while ensuring high accuracy. Specifically, we (1) exploit meta-learning and rehearsal strategies to explicitly cope with data scarcity issues and ensure high accuracy, (2) effectively combine lossless and lossy compression to significantly reduce the resource requirements of CL and rehearsal samples, and (3) developed hardware-aware system on embedded and IoT platforms considering the hardware characteristics. As a result, LifeLearner achieves near-optimal CL performance, falling short by only 2.8% on accuracy compared to an Oracle baseline. With respect to the state-of-the-art (SOTA) Meta CL method, LifeLearner drastically reduces the memory footprint (by 178.7x), end-to-end latency by 80.8-94.2%, and energy consumption by 80.9-94.2%. In addition, we successfully deployed LifeLearner on two edge devices and a microcontroller unit, thereby enabling efficient CL on resource-constrained platforms where it would be impractical to run SOTA methods and the far-reaching deployment of adaptable CL in a ubiquitous manner. Code is available at https://github.com/theyoungkwon/LifeLearner.
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。
