113 research outputs found
Collecting Datasets from Ambient Intelligence Environments
No full textThis paper describes a methodology and lessons learned from collecting datasets in Ambient Intelligence Environments. The authors present considerations on how to setup an experiment and discuss decisions taken at different planning steps, ranging from the selection of human activities over sensor choices to issues of the recording software. The experiment design and execution is illustrated through a dataset involving 150 recording sessions with 28 sensors worn on the subject body and embedded into tools and the environment. The paper also describes a number of unforeseen problems that affected the experiment and useful considerations that help other researchers recording their own ambient intelligence datasets
Benefits of Dynamically Reconfigurable Activity Recognition in Distributed Sensing Environments
No full textThe automatic detection of complex human activities in daily life using distributed ambient and on-body sensors is still an open research challenge. A key issue is to construct scalable systems that can capture the large diversity and variety of human activities. Dynamic system reconfiguration is a possible solution to adaptively focus on the current scene and thus reduce recognition complexity. In this work, we evaluate potential energy savings and performance gains of dynamic reconfiguration in a case study using 28 sensors recording 78 activities performed within four settings. Our results show that reconfiguration improves recognition performance by up to 11.48 %, while reducing energy consumption when turning off unneeded sensors by 74.8 %. The granularity of reconfiguration trades off recognition performance for energy savings
Activity recognition from on-body sensors: accuracy-power trade-off by dynamic sensor selection
No full textActivity recognition from an on-body sensor network enables context-aware applications in wearable computing. A guaranteed classification accuracy is desirable while optimizing power consumption to ensure the system’s wearability. In this paper, we investigate the benefits of dynamic sensor selection in order to use efficiently available energy while achieving a desired activity recognition accuracy. For this purpose we introduce and characterize an activity recognition method with an underlying run-time sensor selection scheme. The system relies on a meta-classifier that fuses the information of classifiers operating on individual sensors. Sensors are selected according to their contribution to classification accuracy as assessed during system training. We test this system by recognizing manipulative activities of assembly-line workers in a car production environment. Results show that the system’s lifetime can be significantly extended while keeping high recognition accuracies. We discuss how this approach can be implemented in a dynamic sensor network by using the context-recognition framework Titan that we are developing for dynamic and heterogeneous sensor networks
Halo modelling in chameleon theories
No full textWe analyse modelling techniques for the large-scale structure formed in scalar-tensor theories of constant Brans-Dicke parameter which match the concordance model background expansion history and produce a chameleon suppression of the gravitational modification in high-density regions. Thereby, we use a mass and environment dependent chameleon spherical collapse model, the Sheth-Tormen halo mass function and linear halo bias, the Navarro-Frenk-White halo density profile, and the halo model. Furthermore, using the spherical collapse model, we extrapolate a chameleon mass-concentration scaling relation from a ΛCDM prescription calibrated to N-body simulations. We also provide constraints on the model parameters to ensure viability on local scales. We test our description of the halo mass function and nonlinear matter power spectrum against the respective observables extracted from large-volume and high-resolution N-body simulations in the limiting case of f(R) gravity, corresponding to a vanishing Brans-Dicke parameter. We find good agreement between the two; the halo model provides a good qualitative description of the shape of the relative enhancement of the f(R) matter power spectrum with respect to ΛCDM caused by the extra attractive gravitational force but fails to recover the correct amplitude. Introducing an effective linear power spectrum in the computation of the two-halo term to account for an underestimation of the chameleon suppression at intermediate scales in our approach, we accurately reproduce the measurements from the N-body simulations
Activity recognition from on-body sensors: accuracy-power trade-off by dynamic sensor selection
No full textActivity recognition from an on-body sensor network enables
context-aware applications in wearable computing. A guaranteed classification
accuracy is desirable while optimizing power consumption to ensure
the system’s wearability. In this paper, we investigate the benefits
of dynamic sensor selection in order to use efficiently available energy
while achieving a desired activity recognition accuracy. For this purpose
we introduce and characterize an activity recognition method with
an underlying run-time sensor selection scheme. The system relies on a
meta-classifier that fuses the information of classifiers operating on individual
sensors. Sensors are selected according to their contribution to
classification accuracy as assessed during system training. We test this
system by recognizing manipulative activities of assembly-line workers
in a car production environment. Results show that the system’s lifetime
can be significantly extended while keeping high recognition accuracies.
We discuss how this approach can be implemented in a dynamic sensor
network by using the context-recognition framework Titan that we are
developing for dynamic and heterogeneous sensor networks
Service discovery and composition in body area networks
No full textIn pervasive environments, Body Area Networks (BANs) are characterized by the mobility of their users. BANs can continuously interact with each other, thus enabling the provision of new applications and services at runtime. New complex services can be provided by composing simpler services available on neighbouring network nodes. However, since the topology of BANs is continuously changing due to users' movements, it is unfeasible to specify a-priori all possible configurations under which a given complex service can be composed. In order to address this issue, we introduce a two--layered service discovery and composition architecture, that proactively notifies a distributed service directory with changes in service availability. In order to cope with the network mobility and intermittent connectivity, our approach is to cluster nodes in the sensor network based on their connectivity patterns. We use a multi--agent state machine to recognize the availability of complex services and to provide them. Our solution is validated by a prototype implementation of our architecture, by the study of the statistical model of complex services, and by experimental evaluations
On the road to per cent accuracy - V. The non-linear power spectrum beyond ΛCDM with massive neutrinos and baryonic feedback
No full textIn the context of forthcoming galaxy surveys, to ensure unbiased constraints on cosmology and gravity when using non-linear structure information, per cent-level accuracy is required when modelling the power spectrum. This calls for frameworks that can accurately capture the relevant physical effects, while allowing for deviations from Lambda cold dark matter (ΛCDM). Massive neutrino and baryonic physics are two of the most relevant such effects. We present an integration of the halo model reaction frameworks for massive neutrinos and beyond ΛCDM cosmologies. The integrated halo model reaction, combined with a pseudo-power spectrum modelled by HMCode2020 is then compared against N-body simulations that include both massive neutrinos and an f(R) modification to gravity. We find that the framework is 4 per cent accurate down to at least ≈ 3 h Mpc-1 for a modification to gravity of |fR0| ≤ 10-5 and for the total neutrino mass Mν Σmν ≤ 0.15 eV. We also find that the framework is 4 per cent consistent with EuclidEmulator2 as well as the Bacco emulator for most of the considered νwCDM cosmologies down to at least k ≈ 3 h Mpc-1. Finally, we compare against hydrodynamical simulations employing HMCode2020's baryonic feedback modelling on top of the halo model reaction. For νΛCDM cosmologies, we find 2 per cent accuracy for Mν ≤ 0.48 eV down to at least k ≈ 5h Mpc-1. Similar accuracy is found when comparing to νwCDM hydrodynamical simulations with Mν = 0.06 eV. This offers the first non-linear, theoretically general means of accurately including massive neutrinos for beyond-ΛCDM cosmologies, and further suggests that baryonic, massive neutrino, and dark energy physics can be reliably modelled independently
Recommended from our members
Smart Sensing and Context: Third European Conference, EuroSSC 2008, Zurich, Switzerland, October 29-31, 2008, Proceedings
No full textThis book constitutes the refereed proceedings of the Third European Conference on Smart Sensing and Context, EuroSSC 2008, held in Zurich, Switzerland, October 29-31, 2008.
The 17 revised full papers presented together with one invited paper were carefully reviewed and selected from 70 submissions. The papers are organized in topical sections on smart objects, spatial and human context inference, context processing and quality of context, as well as context-aware interaction and case studies
- …
