WildFit:面向资源受限物联网系统的自主现场模型适应
WildFit: Autonomous In-situ Model Adaptation for Resource-Constrained IoT Systems
AI总结:
WildFit通过自主现场适应框架,利用背景变化更频繁的特性,在资源受限的物联网系统中实现更高效的模型适应。
摘要:
受限于资源的物联网设备越来越多地依赖深度学习模型,然而这些模型在遇到光照、天气和季节变化等域偏移时会经历显著的精度下降。虽然基于云的重新训练可以解决这个问题,但许多物联网部署存在连接性和能源限制,使得传统微调方法不切实际。我们通过野生动物生态学的视角来探索这一挑战,其中相机陷阱必须在季节、天气和栖息地变化的情况下保持物种分类的准确性,而无需可靠的连接性。我们介绍了WildFit,一种自主的现场适应框架,利用关键洞察:背景场景变化比受监控物种的视觉特征更频繁。WildFit结合了背景感知合成来在设备上生成训练样本,以及域偏移感知微调,仅在必要时触发模型更新以节省资源。我们的背景感知合成在效率基线和扩散模型上分别提升了7.3%和3.0%,同时速度快了多个数量级;我们的域偏移感知微调实现了帕累托最优,使用50%更少的更新和1.5%更高的准确性;端到端系统在消耗仅11.2瓦时37天的情况下,比域适应方法高出20-35%。
英文摘要:
Resource-constrained IoT devices increasingly rely on deep learning models, however, these models experience significant accuracy drops due to domain shifts when encountering variations in lighting, weather, and seasonal conditions. While cloud-based retraining can address this issue, many IoT deployments operate with limited connectivity and energy constraints, making traditional fine-tuning approaches impractical. We explore this challenge through the lens of wildlife ecology, where camera traps must maintain accurate species classification across changing seasons, weather, and habitats without reliable connectivity. We introduce WildFit, an autonomous in-situ adaptation framework that leverages the key insight that background scenes change more frequently than the visual characteristics of monitored species. WildFit combines background-aware synthesis to generate training samples on-device with drift-aware fine-tuning that triggers model updates only when necessary to conserve resources. Our background-aware synthesis surpasses efficient baselines by 7.3% and diffusion models by 3.0% while being orders of magnitude faster, our drift-aware fine-tuning achieves Pareto optimality with 50% fewer updates and 1.5% higher accuracy, and the end-to-end system outperforms domain adaptation approaches by 20-35% while consuming only 11.2 Wh over 37 days-enabling battery-powered deployment.
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。