Daily activity data that records individuals' various types of activities in daily life are widely used in many applications such as activity scheduling, activity recommendation, and policymaking. Though with high value, its accessibility is limited due to high collection costs and potential privacy issues. Therefore, simulating human activities to produce massive high-quality data is of great importance to benefit practical applications. However, existing solutions, including rule-based methods with simplified assumptions of human behavior and data-driven methods directly fitting real-world data, both cannot fully qualify for matching reality. In this paper, motivated by the classic psychological theory, Maslow's need theory describing human motivation, we propose a knowledge-driven simulation framework based on generative adversarial imitation learning. To enhance the fidelity and utility of the generated activity data, our core idea is to model the evolution of human needs as the underlying mechanism that drives activity generation in the simulation model. Specifically, this is achieved by a hierarchical model structure that disentangles different need levels, and the use of neural stochastic differential equations that successfully captures piecewise-continuous characteristics of need dynamics. Extensive experiments demonstrate that our framework outperforms the state-of-the-art baselines in terms of data fidelity and utility. Besides, we present the insightful interpretability of the need modeling. The code is available at https://github.com/tsinghua-fib-lab/SAND.
Energy Expenditure Estimation Through Daily Activity Recognition Using a Smart-phone
University of Paris-Sud(巴黎南大学);CNRS-LIMSI(法国国家科学研究中心-计算机科学与机械与信息学实验室)
本文提出了一个3步系统,估计一个人的实时能量消耗的非侵入性的方式。首先,使用用户的智能手机的传感器,我们建立一个决策树模型来识别他的身体活动(\textit{running},\textit{standing},.)。然后,我们使用检测到的身体活动、时间和用户的速度来推断他的日常活动(\textit{watching TV}、\textit{going to the bathroom}、.)通过使用强化学习环境,部分可观察马尔可夫决策过程框架。一旦日常活动被识别,我们就使用身体活动纲要将这些信息转化为能量消耗。通过在90%的成功检测到8种身体活动,我们在识别17种不同的日常活动时达到了80%的总体准确率。该结果导致我们估计用户的能量消耗,平均误差为预期估计的26%。
Multimodal video and IMU kinematic dataset on daily life activities using affordable devices (VIDIMU)
University of Valladolid(巴利亚多利德大学);University of Applied Sciences and Arts Western Switzerland (HES-SO) Valais-Wallis(瑞士西部应用科学与艺术大学 (HES-SO) 瓦莱-沃利斯分校);University of Geneva(日内瓦大学)
An Efficient Data Imputation Technique for Human Activity Recognition
Instituto de Telecomunicações(电信研究所);Universidade da Beira Interior(贝拉内华大学);Polytechnic Institute of Viseu(维塞乌理工学院);Department of Computer Engineering(计算机工程系);University of Engineering and Technology (UET)(工程与技术大学);Faculty of Computer Science and Engineering(计算机科学与工程学院);University Ss Cyril and Methodius(圣西里尔与美多德大学)
Paving the Way Towards Kinematic Assessment Using Monocular Video: A Preclinical Benchmark of State-of-the-Art Deep-Learning-Based 3D Human Pose Estimators Against Inertial Sensors in Daily Living Activities
Detecting Activities of Daily Living and Routine Behaviours in Dementia Patients Living Alone Using Smart Meter Load Disaggregation
人口老龄化的出现是一个重大的公共卫生问题。这导致患有进行性神经退行性疾病(如痴呆症)的人数增加。因此,这对卫生和社会保健服务造成的压力意味着提供24小时监测是不可持续的。正在考虑技术干预,但是没有解决方案可以非侵入性地监测痴呆症患者的独立生活需求。因此,许多患者在提供干预和支持之前就达到了危机点。与此同时,病人护理依赖于非正式护理人员对重大行为变化的反馈。然而,并非所有人都有社会支持网络,而且痴呆症护理的早期干预往往被错过。智能电表的推出有可能改变这一点。使用机器学习和信号处理技术,家庭能源供应可以被分解以检测哪些家用电器被打开和关闭。这将使日常生活活动(ADL)得到评估,如饮食,并观察到日常生活的变化,以进行早期干预。主要目的是帮助减少病情恶化,使患者能够在家中呆更长时间。一个支持向量机(SVM)和随机决策森林分类器建模使用的数据从三个测试家庭。然后,在与Mersey Care NHS Foundation Trust合作进行的为期六个月的临床试验中,使用经过训练的模型来监测两名痴呆症患者。在用于电器检测的负载分解的情况下,实现了SVM(AUC=0.86074,Sen=0.756和Spec=0.92838)。而决策森林达到(AUC=0.9429,Sen=0.9634和Spec=0.9634)。还分析了ADL,以识别居住者的行为模式,同时检测日常生活的改变。
InterAct: A Large-Scale Dataset of Dynamic, Expressive and Interactive Activities between Two People in Daily Scenarios
The University of Hong Kong(香港大学);Centre for Transformative Garment Production(转型服装生产中心);Great Bay University(大湾大学);Dongguan Key Laboratory for Intelligence and Information Technology(东莞智能与信息科技重点实验室);Shandong University(山东大学);National Institute of Informatics(国家信息研究所)
CHARM: A Hierarchical Deep Learning Model for Classification of Complex Human Activities Using Motion Sensors
Meta Platforms, Inc.(元平台公司)
在本文中,我们报告了一个层次化的深度学习模型,用于使用运动传感器对复杂的人类活动进行分类。与用于基于事件的活动识别(例如计步、跌倒检测和手势识别)的传统人类活动识别(HAR)模型相比,我们将这种新的深度学习模型称为CHARM。复杂人类活动识别模型(Complex Human Activity Recognition Model),旨在识别由多个不同的低级别活动以非确定性顺序组成的高级别人类活动,例如准备膳食、家务和日常事务。CHARM不仅在平均准确度和F1分数方面在数量上优于高级活动识别的最先进的监督学习方法,而且还自动学习识别低级活动,例如操作手势和运动模式,而无需对此类活动进行任何显式标记。这为使用可穿戴传感器的人机交互(HMI)模态开辟了新的途径,其中用户可以选择将自动化任务与高级活动相关联,诸如控制家庭自动化(例如,机器人真空吸尘器、灯和恒温器)或在正确的时间呈现上下文相关的信息(例如,提醒、状态更新和天气/新闻报道)。此外,当仅使用高级活动标签进行训练时学习低级用户活动的能力可以为固有地难以标记的HAR任务的半监督学习铺平道路。
RF Sensing for Continuous Monitoring of Human Activities for Home Consumer Applications
Center for Advanced Communication, Villanova University(维拉诺瓦大学先进通信中心);Comcast Labs, Comcast(康卡斯特实验室(康卡斯特公司));Villanova University(维拉诺瓦大学);Comcast(康卡斯特公司)
NOIR 2.0: Neural Signal Operated Intelligent Robots for Everyday Activities
Institute for Computational and Mathematical Engineering(计算与数学工程研究所);Department of Computer Science(计算机科学系);Department of Mechanical Engineering(机械工程系)