In this study, a novel deep neural network (DNN) architecture, CarSpeedNet,
is introduced to estimate car speed using three-axis accelerometer data from
smartphones. Utilizing 13 hours of data collected from smartphones mounted in
vehicles navigating through various regions in Israel, the CarSpeedNet
effectively learns the relationship between measured smartphone acceleration
and car speed. Ground truth speed data was obtained at 1[Hz] from the GPS
receiver in the smartphones. The proposed model enables high-frequency speed
estimation, incorporating historical inputs. Our trained model demonstrates
exceptional accuracy in car speed estimation, achieving a precision of less
than 0.72[m/s] during an extended driving test, solely relying on smartphone
accelerometer data without any connectivity to the car.

通过使用智能手机的三轴加速度计数据，本研究引入了一种新颖的深度神经网络（DNN）架构 CarSpeedNet，用于估计汽车速度。通过收集在以色列不同地区行驶的车辆中安装的智能手机的 13 小时数据，CarSpeedNet 有效地学习了测量的智能手机加速度和汽车速度之间的关系。作者提出的模型能够进行高频速度估计，并整合历史输入。在仅依靠智能手机加速度计数据而不与汽车进行任何连接的情况下，在一次长时间驾驶测试中，我们训练的模型表现出了卓越的车速估计准确性，达到了小于 0.72 [m/s] 的精度。

CarSpeedNet：基于智能手机加速度计的深度神经网络汽车速度估计

CarSpeedNet: A Deep Neural Network-based Car Speed Estimation from  Smartphone Accelerometer

Human Activity Recognition (HAR) enables context-aware user experiences where
mobile apps can alter content and interactions depending on user activities.
Hence, smartphones have become valuable for HAR as they allow large, and
diversified data collection. Although previous work in HAR managed to detect
simple activities (i.e., sitting, walking, running) with good accuracy using
inertial sensors (i.e., accelerometer), the recognition of complex daily
activities remains an open problem, specially in remote work/study settings
when people are more sedentary. Moreover, understanding the everyday activities
of a person can support the creation of applications that aim to support their
well-being. This paper investigates the recognition of complex activities
exclusively using smartphone accelerometer data. We used a large smartphone
sensing dataset collected from over 600 users in five countries during the
pandemic and showed that deep learning-based, binary classification of eight
complex activities (sleeping, eating, watching videos, online communication,
attending a lecture, sports, shopping, studying) can be achieved with AUROC
scores up to 0.76 with partially personalized models. This shows encouraging
signs toward assessing complex activities only using phone accelerometer data
in the post-pandemic world.

本研究使用手机加速度计数据对 8 种复杂活动进行基于深度学习的二分类，取得了 AUROC 得分高达 0.76 的良好结果，表明在后疫情时代使用手机加速度计数据对复杂活动进行识别的可行性。