Eric Guenterberg

Sport training using body sensor networks: a statistical approach to measure wrist rotation for golf swing

Athletes in any sports can greatly benefit from feedback systems for improving the quality of their training. In this paper, we present a golf swing training system which incorporates wearable motion sensors to obtain inertial information and provide feedback on the quality of movements. The sensors are placed on a golf club and athletes body at positions which capture the unique movements of a golf swing. We introduce a quantitative model which takes into consideration signal processing techniques on the collected data and quantifies the correctness of the performed actions. We evaluate the effectiveness of our framework on data obtained from four subjects and discuss ongoing research.

Energy-Efficient Information-Driven Coverage for Physical Movement Monitoring in Body Sensor Networks

Advances in technology have led to the development of various light-weight sensor devices that can be woven into the physical environment of our daily lives. Such systems enable on-body and mobile health-care monitoring. Our interest particularly lies in the area of movement-monitoring platforms that operate with inertial sensors. In this paper, we introduce the notion of compatibility graphs and describe how they can be utilized for power optimization. We first formulate an action coverage problem that will consider the sensing coverage from a collaborative signal processing perspective. Our solution is capable of eliminating redundant sensor nodes while maintaining the quality of service. The problem we outline can be transformed into an NP-hard problem. Therefore, we propose an ILP formulation to attain a lower bound on the solution and a fast greedy technique. Moreover, we present a system for dynamically activating and deactivating sensor nodes in real time. We then use our graph representation to develop an efficient formulation for maximum lifetime. This formulation provides sufficient information for finding activation duties for each sensor node. Finally, we demonstrate the effectiveness of our techniques on data collected from several subjects.

A phonological expression for physical movement monitoring in body sensor networks

Monitoring human activities using wearable wireless sensor nodes has the potential to enable many useful applications for everyday situations. The deployment of a compact and computationally efficient grammatical representation of actions reduces the complexities involved in the detection and recognition of human behaviors in a distributed system. In this paper, we introduce a road map to a linguistic framework for the symbolic representation of inertial information for physical movement monitoring. Our method for creating phonetic descriptions consists of constructing primitives across the network and assigning certain primitives to each movement. Our technique exploits the notion of a decision tree to identify atomic actions corresponding to every given movement. We pose an optimization problem for the fast identification of primitives. We then prove that this problem is NP-Complete and provide a fast greedy algorithm to approximate the solution. Finally, we demonstrate the effectiveness of our phonetic model on data collected from three subjects.

Action coverage formulation for power optimization in body sensor networks

Advances in technology have led to the development of various light-weight sensory devices that can be woven into the physical environment of our daily lives. Such systems enable on-body and mobile health-care monitoring. Our interest particularly lies in the area of movement monitoring platforms that operate with inertial sensors. In this paper, we propose a power optimization technique that will consider the sensing coverage problem from a collaborative signal processing perspective. We introduce compatibility graphs and describe how they can be utilized for power optimization. The problem we outline can be transformed into an NP-hard problem. Therefore, we propose an ILP formulation to attain a lower bound on the solution and a fast greedy technique. Along side this, we introduce a system for dynamically activating and deactivating sensor nodes in real-time. Finally, we elucidate the effectiveness of our techniques on data collected from several subjects.

A Method for Extracting Temporal Parameters Based on Hidden Markov Models in Body Sensor Networks With Inertial Sensors

Human movement models often divide movements into parts. In walking, the stride can be segmented into four different parts, and in golf and other sports, the swing is divided into sections based on the primary direction of motion. These parts are often divided based on key events, also called temporal parameters. When analyzing a movement, it is important to correctly locate these key events, and so automated techniques are needed. There exist many methods for dividing specific actions using data from specific sensors, but for new sensors or sensing positions, new techniques must be developed. We introduce a generic method for temporal parameter extraction called the hidden Markov event model based on hidden Markov models. Our method constrains the state structure to facilitate precise location of key events. This method can be quickly adapted to new movements and new sensors/sensor placements. Furthermore, it generalizes well to subjects not used for training. A multiobjective optimization technique using genetic algorithms is applied to decrease error and increase cross-subject generalizability. Further, collaborative techniques are explored. We validate this method on a walking dataset by using inertial sensors placed on various locations on a human body. Our technique is designed to be computationally complex for training, but computationally simple at runtime to allow deployment on resource-constrained sensor nodes.

Wireless Medical-Embedded Systems: A Review of Signal-Processing Techniques for Classification

Body-worn sensor systems will help to revolutionize the medical field by providing a source of continuously collected patient data. This data can be used to develop and track plans for improving health (more sleep and exercise), detect disease early, and provide an alert for dangerous events (e.g., falls and heart attacks). The amount of data collected by even a small set of sensors running all day is too much for any person to analyze. Signal processing and classification can be used to automatically extract useful information. This paper presents a general classification framework for wireless medical devices and reviews the available literature for signal processing and classification systems or components used in body-worn sensor systems. Examples focus on electrocardiography classification and signal processing for inertial sensors.

Distributed Continuous Action Recognition Using a Hidden Markov Model in Body Sensor Networks

One important application of Body Sensor Networks is action recognition. Action recognition often implicitly requires partitioning the sensor data into intervals, then labeling the partitions according to the actions each represents or as a non-action. The temporal partitioning stage is called segmentation and the labeling is called classification. While many effective methods exist for classification, segmentation remains problematic. We present a technique inspired by continuous speech recognition that combines segmentation and classification using Hidden Markov Models. This technique is distributed and only involves limited data sharing between sensor nodes. We show the results of this technique and the bandwidth savings over full data transmission.

A Distributed Hidden Markov Model for Fine-grained Annotation in Body Sensor Networks

Human movement models often divide movements into parts. In walking the stride can be segmented into four different parts, and in golf and other sports, the swing is divided into section based on the primary direction of motion. When analyzing a movement, it is important to correctly locate the key events dividing portions. There exist methods for dividing certain actions using data from speci¿c sensors. We introduce a generalized method for event annotation based on Hidden Markov Models. Genetic algorithms are used for feature selection and model parameterization. Further, collaborative techniques are explored. We validate this method on a walking dataset using inertial sensors placed on various locations on a human body. Our technique is computationally simple to allow it to run on resource constrained sensor nodes.

An automatic segmentation technique in body sensor networks based on signal energy

Monitoring human activities using wearable wireless sensor nodes has the potential to enable many useful applications for everyday situations. The long-term lifestyle monitoring can greatly improve healthcare by gathering information about quality of life; aiding the diagnosis and tracking of certain diseases such as Parkinsons. The deployment of an automatic and computationally-efficient algorithm reduces the complexities involved in the detection and recognition of human activities in a distributed system. This paper presents a new algorithm for automatic segmentation of routine human activities. The proposed algorithm can distinguish between discrete periods of activity and rest without specifically knowing the activity. A finite subset of nodes can detect all human activities, but each node by itself can only detect a particular set of activities. For local segmentation we choose the parameters for each node that result in the least segmentation error. We demonstrate the effectiveness of our algorithm on data collected from body sensor networks for a scenario simulating a set of daily activities.

Human identification by gait analysis

Human movement monitoring using wireless sensors has become an important area of research today. The use of wireless sensors in human identification is a relatively new idea with interesting applications in portable device security and user recognition. In this paper, we describe a real-time wireless sensor system based on inexpensive inertial sensors that uses gait analysis to uniquely identify subjects.