Supplementary Data for "CLOAR: CSI-based Long-term Occupant Activity Recognition"

With the rapid deployment of indoor Wi-Fi networks, Channel State Information (CSI) has been used for device-free occupant activity recognition. However, various environmental factors interfere with the stable propagation of Wi-Fi signals indoors, which causes temporal variation of CSI data. In this study, we investigated temporal CSI variation in a real-world housing environment and its impact on learning-based occupant activity recognition. The CSI variation over time changes distributions of the CSI data, and the pre-trained model’s accuracy performance becomes degraded during long-term monitoring. In order to address the temporal dependency issue, we developed an effective long-term occupant activity recognition model based on the semi-supervised meta-learning approach. Our model leveraged unlabeled target data with its pseudo labels and synthesized numerous query datasets using mixup-based data augmentation, which generalized the model during training. The model provided an average of 91.09% activity classification accuracy for the target data, which had different statistical characteristics from the source data. This result demonstrates that our model can reliably monitor occupant activities for long-term periods.

随着室内 Wi-Fi 网络的快速部署，信道状态信息（Channel State Information，简称 CSI）已被应用于无设备占用者活动识别。然而，诸多环境因素干扰了室内 Wi-Fi 信号的稳定传播，导致 CSI 数据的时变性。在本研究中，我们调查了现实住宅环境中 CSI 的时变特性及其对基于学习的占用者活动识别的影响。随着时间的推移，CSI 的变化改变了 CSI 数据的分布，使得预训练模型在长期监测期间准确性表现下降。为了解决时序依赖性问题，我们基于半监督元学习策略开发了一种有效的长期占用者活动识别模型。我们的模型利用带伪标签的无标签目标数据，并使用基于 mixup 的数据增强技术合成多个查询数据集，以泛化模型在训练过程中的表现。对于具有与源数据不同统计特性的目标数据，该模型实现了平均 91.09% 的活动分类准确率。这一结果证明了我们的模型能够可靠地监测占用者的长期活动。