@article {seiter2014rehabilitation, title = {Daily Life Activity Routine Discovery in Hemiparetic Rehabilitation Patients Using Topic Models}, journal = {Methods of Information in Medicine}, year = {In Press}, issn = {1574-1192}, author = {Julia Seiter and Adrian Derungs and Corina Schuster-Amft and Oliver Amft and Gerhard Tr{\"o}ster} } @article {seiter2014discovery, title = {Discovery of activity composites using topic models: An analysis of unsupervised methods}, journal = {Pervasive and Mobile Computing}, volume = {15}, year = {2014}, note = {Special Issue on Data Mining in Pervasive Environments}, pages = {215 - 227}, abstract = {Abstract In this work we investigate unsupervised activity discovery approaches using three topic model~(TM) approaches, based on Latent Dirichlet Allocation~(LDA), n -gram TM~(NTM), and correlated TM~(CTM). While \{LDA\} structures activity primitives, \{NTM\} adds primitive sequence information, and \{CTM\} exploits co-occurring topics. We use an activity composite/primitive abstraction and analyze three public datasets with different properties that affect the discovery, including primitive rate, activity composite specificity, primitive sequence similarity, and composite-instance ratio. We compare the activity composite discovery performance among the \{TM\} approaches and against a baseline using k -means clustering. We provide guidelines for method and optimal \{TM\} parameter selection, depending on data properties and activity primitive noise. Results indicate that \{TMs\} can outperform k -means clustering up to 17\%, when composite specificity is low. LDA-based \{TMs\} showed higher robustness against noise compared to other \{TMs\} and k -means.}, keywords = {Activity routines}, issn = {1574-1192}, doi = {http://dx.doi.org/10.1016/j.pmcj.2014.05.007}, url = {http://www.sciencedirect.com/science/article/pii/S1574119214000832}, author = {Julia Seiter and Oliver Amft and Mirco Rossi and Gerhard Tr{\"o}ster} } @conference {Derungs2014-P_ACOMORE, title = {Motion-adaptive Duty-cycling to Estimate Orientation using Inertial Sensors}, booktitle = {ACOMORE 2014: IEEE International Conference on Pervasive Computing and Communications Workshops}, series = {PerCom Workshops}, year = {2014}, note = {1st Symposium on Activity and Context Modeling and Recognition}, pages = {47{\textendash}54}, publisher = {IEEE}, organization = {IEEE}, abstract = {

We present a motion-adaptive duty-cycling approach to estimate orientation using inertial sensors. In particular, we deploy a proportional forward-controller to adjust the duty-cycle of inertial sensing units\ (IMU) and the orientation estimation update rate of an extended Kalman filter\ (EKF). In sample data recordings and a simulated daily life dataset from a wrist-worn IMU, we show that our motion-adaptive approach incurs substantially lower errors that a static duty-cycling approach. During phases with low or no rotation motion, as it is often occurring in daily activities, our approach can dynamically reduce the IMU operation to 20\% of the regular rate. Results show that duty-cycles of 50\% are common during low-wrist rotation activities, such as reading and typing, while orientation error is below 1$\degree$. We further show the power saving benefits of our approach in a case study of the ETHOS IMU device.

}, doi = {10.1109/PerComW.2014.6815163}, author = {Adrian Derungs and Han Lin and Holger Harms and Oliver Amft} } @inbook {196, title = {1st Workshop on Human Factors and Activity Recognition in Healthcare, Wellness and Assisted Living (Recognise2Interact)}, booktitle = {UbiComp{\textquoteright}13 ACM International Joint Conference on Pervasive and Ubiquitous Computing}, year = {2013}, publisher = {Association for Computing Machinery}, organization = {Association for Computing Machinery}, isbn = {978-1-4503-2139-6/13/09}, author = {Pierluigi Casale and Steven Houben and Oliver Amft} } @conference {204, title = {COPDTrainer: A Smartphone-based Motion Rehabilitation Training System with Real-Time Acoustic Feedback}, booktitle = {Ubicomp 2013: Proceedings of the 2013 ACM International Joint Conference on Pervasive and Ubiquitous Computing}, year = {2013}, publisher = {ACM}, organization = {ACM}, abstract = {

Patient motion training requires adaptive, personalized exercise models and systems that are easy to handle. In this paper, we evaluate a training system based on a smartphone that integrates in clinical routines and serves as a tool for therapist and patient. Only the smartphone?s build-in inertial sensors were used to monitor exercise execution and providing acoustic feedback on exercise performance and exercise errors. We used a sinusoidal motion model to exploit the typical repetitive structure of motion exercises. A Teach-mode was used to personalize the system by training under the guidance of a therapist and deriving exercise model parameters. Subsequently, in a Train-mode, the system provides exercise feedback. We validate our approach in a validation with healthy volunteers and in an intervention study with COPD patients. System performance, trainee performance, and feedback efficacy were analysed. We further compare the therapist and training system performances and demonstrate that our approach is viable.

}, doi = {10.1145/2493432.2493454}, author = {Gabriele Spina and Guannan Huang and Anouk W. Vaes and Martijn A. Spruit and Oliver Amft} } @conference {203, title = {CRNTC+: A Smartphone-based Sensor Processing Framework for Prototyping Personal Healthcare Applications}, booktitle = {PH 2013: Proceedings of the 7th International Conference on Pervasive Computing Technologies for Healthcare}, year = {2013}, publisher = {IEEE}, organization = {IEEE}, abstract = {

While smartphone apps for health monitoring and patient support are of great interest to care providers and patients alike, suitable development and evaluation frameworks are currently lacking. We present and evaluate an Android open-source smartphone framework CRNTC+ for sensors data acquisition, signal processing, pattern analysis, interaction and feedback, based on the Context Recognition Network Toolbox\ (CRNT). CRNTC+ extends the original CRNT by providing components to read smartphone and external sensor data, supporting annotations, and various output components. Here, we formally evaluate CRNTC+ regarding extensibility, scalability, and energy consumption. We present study results where CRNTC+ was deployed in an application to detect epileptic seizures. Results showed that CRNTC+ is well-suited for prototyping health applications in real-life, where online sensor data recording and recognition is needed.

}, author = {Gabriele Spina and Frank Roberts and Jens Weppner and Paul Lukowicz and Oliver Amft} } @conference {DBLP:conf/percom/SmithDTAZ13, title = {Exploring Concept Drift using Interactive Simulations}, booktitle = {PerCom Workshops}, year = {2013}, pages = {49-54}, author = {Jeremiah Smith and Naranker Dulay and M{\'a}t{\'e} Attila T{\'o}th and Oliver Amft and Yanxia Zhang} } @conference {Casale2013-P_AMI, title = {Inferring Model Structures from Inertial Sensor Data in Distributed Activity Recognition Scenarios}, booktitle = {AMI 2013: Proceedings of the Joint Conference on Ambient Intelligence}, year = {2013}, pages = {62{\textendash}77}, publisher = {Springer}, organization = {Springer}, abstract = {

Activity-Events-Detectors digraphs describe the relations between human activities and sensor nodes under a distributed perspective. The graphs provide a conceptual abstraction that decouples the set of activities from the sensor network with the aim of improving the recognition performances and lowering the computational constraints of the detection tasks in the sensor node. In this work, a data-driven methodology that learns groups of activities and infers the con?guration of detectors embedded in the sensors nodes of the network is proposed. The methodology, de?ned on a clustering procedure, derives and infers all the relevant information from sensors data, making no a-priori assumptions on the relations between sensor nodes and activities. Using the inferred structured models, a performance boost of 15\% in the ?nal classi?cation accuracy is obtained with a signi?cant reduction of the computational resources needed for recognition purposes.

}, keywords = {clustering, context recognition, inertial sensors, wireless sensor networks}, author = {Pierluigi Casale and Oliver Amft} } @conference {201, title = {Personalizing Energy Expenditure Estimation Using a Cardiorespiratory Fitness Predicate}, booktitle = {PH 2013: Proceedings of the 7th International Conference on Pervasive Computing Technologies for Healthcare}, year = {2013}, publisher = {IEEE}, organization = {IEEE}, abstract = {

Accurate Energy Expenditure (EE) estimation is key in understanding how behavior and daily physical activity (PA) patterns affect health, especially in today\&$\#$39;s sedentary society. Wearable accelerometers (ACC) and heart rate (HR) sensors have been widely used to monitor physical activity and estimate EE. However, current EE estimation algorithms have not taken into account a person\&$\#$39;s cardiorespiratory fitness (CRF), even though CRF is the main cause of inter-individual variation in HR during exercise. In this paper we propose a new algorithm, which is able to significantly reduce EE estimate error and inter-individual variability, by automatically modeling CRF, without requiring users to perform specific fitness tests. Results show a decrease in Root Mean Square Error (RMSE) between 28 and 33\% for walking, running and biking activities, compared to state of the art activity-specific EE algorithms combining ACC and HR.

}, author = {Marco Altini and Julien Penders and Oliver Amft} } @article {200, title = {Smart Textiles: From Niche to Mainstream}, journal = {IEEE Pervasive Computing}, volume = {12}, year = {2013}, month = {July-Sept}, pages = {81{\textendash}84}, abstract = {

Current technology supports only special-purpose, low-volume textiles, garments, and electronics. Moreover, the textile, electronic, and software industries have different product cycles, cultures, and price models, creating scores of practical problems for smart textiles. Mass producing smart cloth will require decoupling the textile production from concrete sensing apps and moving the complexity to generic electronics and software\–creating wearable sensing as an app.

}, doi = {10.1109/MPRV.2013.55}, author = {Jingyuan Cheng and Paul Lukowicz and Niels Henze and Albrecht Schmidt and Oliver Amft and Giovanni A. Salvatore and Gerhard Tr{\"o}ster} } @conference {202, title = {Using RFID Tags as Reference for Phone Location and Orientation in Daily Life}, booktitle = {AH 2013: Proceedings of the 4th International Augmented Human Conference}, year = {2013}, publisher = {ACM}, organization = {ACM}, abstract = {

This paper investigates a novel approach to obtain location and orientation annotation for smartphones in real-life recordings. We attached RFID tags to places where phones are located in daily life, such as pockets and backpacks. The RFID reader integrated in modern smartphones was used to continuously scan for registered tags. In a first evaluation across several full-day recordings and using nine locations, our approach achieved an accuracy of 80\ \% when compared to a manual diary. Only 5.3\ \% of all tags were missed. We conclude that RFID-based location and orientation tagging is a viable option to obtain ground truth reference for real-life activity recognition algorithm developments.

}, keywords = {ground truth, NFC, unsupervised annotation}, doi = {10.1145/2459236.2459269}, author = {Florian Wahl and Oliver Amft} }