一句话：提出首个无需用户反馈的端到端系统，在移动设备上持续监控并自动适配 ML 模型，缓解精度退化。
机器学习模型越来越多地部署在移动设备本地运行，以获得低延迟推理和离线能力。然而，ML 运维人员难以跟踪设备端模型精度，而本地数据漂移等原因会导致精度不可预测地退化。本文提出 Nazar，这是首个无需用户反馈即可在移动设备上持续监控并适配模型的端到端系统。其关键观察是：精度退化往往由特定根因导致，且可能影响大量设备。Nazar 一旦检测到影响大量设备的退化，便自动定位根因，并针对这些根因对模型进行适配。在两个计算机视觉数据集上的评估表明，相比现有方法，Nazar 能持续提升精度，最高提升 19.4%。
ML models are increasingly run locally on mobile devices for low-latency inference and offline operation. However, it is hard for ML operators to track on-device model accuracy, which can degrade unpredictably (e.g. due to local data drift). We design Nazar, the first end-to-end system for continuously monitoring and adapting models on mobile devices without requiring feedback from users. Our key observation is that accuracy degradation is often due to a specific root cause, which may affect a large group of devices. Once Nazar detects a degradation affecting a large number of devices, it automatically pinpoints the root causes and adapts the model specifically to them. Evaluation on two computer vision datasets shows that Nazar consistently boosts accuracy compared to existing approaches by up to 19.4%.
@inproceedings{HaoWHLKDMYC25,
title = {Nazar: Monitoring and Adapting ML Models on Mobile Devices},
author = {Wei Hao and Zixi Wang and Lauren Hong and Lingxiao Li and Nader Karayanni and AnMei Dasbach-Prisk and Chengzhi Mao and Junfeng Yang and Asaf Cidon},
booktitle = {ASPLOS},
year = {2025},
doi = {10.1145/3669940.3707246}
}