Dynamic Adaptation of Opportunistic Sensor Configurations for Continuous and Accurate Activity Recognition
نویسندگان
چکیده
An ever-larger availability of devices that are attached with different sensing capabilities (e.g., smart phones) shifted the challenge in activity and context recognition from the application specific deployment of new sensors to the utilization of already available devices. Therefore, a system that operates in an opportunistic way has to take advantage of the currently available sensing infrastructure in terms of utilizing sensors in form of ensembles that are best suited to execute a specific activity recognition task. Continuous, stable, and accurate activity recognition can be assured if such a system is able to react in real-time to such dynamics in the sensing infrastructure. In detail, this paper tackles the characteristic application cases where sensors spontaneously appear, disappear and reappear in the sensing infrastructure and evaluates the continuousness and stability of the selfadaption methods within the OPPORTUNITY Framework, which is a reference implementation of an opportunistic activity and context recognition system. Keywords-Opportunistic sensing; activity and context recognition; self-adaptation.
منابع مشابه
Sensor Abstractions for Opportunistic Activity and Context Recognition Systems
Pervasive environments are inherently characterized to draw from sensor infrastructures in order to become situation aware. Very recent technological evolutions of sensor hardware (e.g. for geoposition, acceleration, orientation, noise, light, humidity, chemical properties, etc.) have fertilized an explosive growth of sensor infrastructures, introducing whole new challenges for sensor software ...
متن کاملSpeaker Adaptation in Continuous Speech Recognition Using MLLR-Based MAP Estimation
A variety of methods are used for speaker adaptation in speech recognition. In some techniques, such as MAP estimation, only the models with available training data are updated. Hence, large amounts of training data are required in order to have significant recognition improvements. In some others, such as MLLR, where several general transformations are applied to model clusters, the results ar...
متن کاملSpeaker Adaptation in Continuous Speech Recognition Using MLLR-Based MAP Estimation
A variety of methods are used for speaker adaptation in speech recognition. In some techniques, such as MAP estimation, only the models with available training data are updated. Hence, large amounts of training data are required in order to have significant recognition improvements. In some others, such as MLLR, where several general transformations are applied to model clusters, the results ar...
متن کاملTowards a Framework for opportunistic Activity and Context Recognition
Due to recent advancements in the development of sensor devices which results in increasing mobility and computing power and decreasing size and costs, this paper presents two features for a sensor-based opportunistic activity and context recognition enabling framework. Opportunistic in this context means that no sensor infrastructure has to be predefined at the design time of the system, it ra...
متن کاملSeventh Framework Programme
sensor independent features. Different physical quantities can provide the same abstract information about an activity. Thus, for example, on-body inertial sensors, clothing-integrated textile elongation sensors, and visual tracking, all give information about body parts trajectories. If the classifiers are trained on such trajectories rather than on raw sensor signals, then the classification ...
متن کامل