Physical activity recognition from accelerometer data using a multi-scale ensemble method

Accurate and detailed measurement of an individual's physical activity is a key requirement for helping researchers understand the relationship between physical activity and health. Accelerometers have become the method of choice for measuring physical activity due to their small size, low cost, convenience and their ability to provide objective information about physical activity. However, interpreting accelerometer data once it has been collected can be challenging. In this work, we applied machine learning algorithms to the task of physical activity recognition from triaxial accelerometer data. We employed a simple but effective approach of dividing the accelerometer data into short non-overlapping windows, converting each window into a feature vector, and treating each feature vector as an i.i.d training instance for a supervised learning algorithm. In addition, we improved on this simple approach with a multi-scale ensemble method that did not need to commit to a single window size and was able to leverage the fact that physical activities produced time series with repetitive patterns and discriminative features for physical activity occurred at different temporal scales.

The use of uniaxial and triaxial accelerometers to measure children's "free-play" physical activity

In order to effectively measure the physical activity of children, objective monitoring devices must be able to quantify the intermittent and nonlinear movement of free play. The purpose of this study was to investigate the validity of the Computer Science and Applications (CSA) uniaxial accelerometer and the TriTrac-R3D triaxial accelerometer with respect to their ability to measure 8 "free-play" activities of different intensity. The activities ranged from light to very vigorous in intensity and included activities such as throwing and catching, hopscotch, and basketball. Twenty-eight children, ages 9 to 11, wore a CSA and a heart rate monitor while performing the activities. Sixteen children also wore a Tritrac. Counts from the CSA, Tritrac, and heart rates corresponding to the last 3 min of the 5 min spent at each activity were averaged and used in correlation analyses. Across all 8 activities, Tritrac counts were significantly correlated with predicted MET level (r= 0.69) and heart rate (r= 0.73). Correlations between CSA output, predicted MET level (0.43), and heart rate (0.64) were also significant but were lower than those observed for the Tritrac. These data indicate that accelerometers are an appropriate methodology for measuring children's free-play physical activities.

Machine learning for activity recognition : hip versus wrist data

Problem addressed Wrist-worn accelerometers are associated with greater compliance. However, validated algorithms for predicting activity type from wrist-worn accelerometer data are lacking. This study compared the activity recognition rates of an activity classifier trained on acceleration signal collected on the wrist and hip. Methodology 52 children and adolescents (mean age 13.7 +/- 3.1 year) completed 12 activity trials that were categorized into 7 activity classes: lying down, sitting, standing, walking, running, basketball, and dancing. During each trial, participants wore an ActiGraph GT3X+ tri-axial accelerometer on the right hip and the non-dominant wrist. Features were extracted from 10-s windows and inputted into a regularized logistic regression model using R (Glmnet + L1). Results Classification accuracy for the hip and wrist was 91.0% +/- 3.1% and 88.4% +/- 3.0%, respectively. The hip model exhibited excellent classification accuracy for sitting (91.3%), standing (95.8%), walking (95.8%), and running (96.8%); acceptable classification accuracy for lying down (88.3%) and basketball (81.9%); and modest accuracy for dance (64.1%). The wrist model exhibited excellent classification accuracy for sitting (93.0%), standing (91.7%), and walking (95.8%); acceptable classification accuracy for basketball (86.0%); and modest accuracy for running (78.8%), lying down (74.6%) and dance (69.4%). Potential Impact Both the hip and wrist algorithms achieved acceptable classification accuracy, allowing researchers to use either placement for activity recognition.

Engaging in videogame play: An activity-centric analysis of the player experience

This paper focuses on examining play activities in people's favourite videogame experience. Through interviews with 30 videogame players we discovered which types of play activities are most appealing. Our research identifies the level of appeal of a wide range of game play activities. We have established that high levels of engagement for many participants is grounded in play as power and play as strategy, with play as fantasy adding to the experience. Through our study we established that conflict-based activities hold strong appeal. We subsequently investigated the context in which players talked about their experience of conflict within game. By using activities as a categorisation of gameplay we have been able to capture the play experience across a range of games and a range of gaming contexts. By examining players' individual experience we begin to understand why conflict in videogames appears to be a popular choice of activities.

Activity recognition using binary tree SVM

This paper presents an effective classification method based on Support Vector Machines (SVM) in the context of activity recognition. Local features that capture both spatial and temporal information in activity videos have made significant progress recently. Efficient and effective features, feature representation and classification plays a crucial role in activity recognition. For classification, SVMs are popularly used because of their simplicity and efficiency; however the common multi-class SVM approaches applied suffer from limitations including having easily confused classes and been computationally inefficient. We propose using a binary tree SVM to address the shortcomings of multi-class SVMs in activity recognition. We proposed constructing a binary tree using Gaussian Mixture Models (GMM), where activities are repeatedly allocated to subnodes until every new created node contains only one activity. Then, for each internal node a separate SVM is learned to classify activities, which significantly reduces the training time and increases the speed of testing compared to popular the `one-against-the-rest' multi-class SVM classifier. Experiments carried out on the challenging and complex Hollywood dataset demonstrates comparable performance over the baseline bag-of-features method.

Elliptical summary randomisation for sensor-based human activity recognition

Many conventional statistical machine learning al- gorithms generalise poorly if distribution bias ex- ists in the datasets. For example, distribution bias arises in the context of domain generalisation, where knowledge acquired from multiple source domains need to be used in a previously unseen target domains. We propose Elliptical Summary Randomisation (ESRand), an efficient domain generalisation approach that comprises of a randomised kernel and elliptical data summarisation. ESRand learns a domain interdependent projection to a la- tent subspace that minimises the existing biases to the data while maintaining the functional relationship between domains. In the latent subspace, ellipsoidal summaries replace the samples to enhance the generalisation by further removing bias and noise in the data. Moreover, the summarisation enables large-scale data processing by significantly reducing the size of the data. Through comprehensive analysis, we show that our subspace-based approach outperforms state-of-the-art results on several activity recognition benchmark datasets, while keeping the computational complexity significantly low.

Google Android : a comprehensive introduction

Google Android, Google's new product and its first attempt to enter the mobile market, might have an equal impact on mobile users like Apple's hyped product, the iPhone. In this Technical report we are going to present the Google Android platform, what Android is, describe why it might be considered as a worthy rival to Apple's iPhone. We will describe parts of its internals, take a look "under the hood" while explaining components of the underlying operating system. We will show how to develop applications for this platform, which difficulties a developer might have to face, and how developers can possibly use other programming languages to develop for Android than the propagated language Java.

Multiple instance dictionary learning for activity representation

This paper presents an effective feature representation method in the context of activity recognition. Efficient and effective feature representation plays a crucial role not only in activity recognition, but also in a wide range of applications such as motion analysis, tracking, 3D scene understanding etc. In the context of activity recognition, local features are increasingly popular for representing videos because of their simplicity and efficiency. While they achieve state-of-the-art performance with low computational requirements, their performance is still limited for real world applications due to a lack of contextual information and models not being tailored to specific activities. We propose a new activity representation framework to address the shortcomings of the popular, but simple bag-of-words approach. In our framework, first multiple instance SVM (mi-SVM) is used to identify positive features for each action category and the k-means algorithm is used to generate a codebook. Then locality-constrained linear coding is used to encode the features into the generated codebook, followed by spatio-temporal pyramid pooling to convey the spatio-temporal statistics. Finally, an SVM is used to classify the videos. Experiments carried out on two popular datasets with varying complexity demonstrate significant performance improvement over the base-line bag-of-feature method.

Semi-supervised intelligent surveillance system for secure environments

This paper proposes a semi-supervised intelligent visual surveillance system to exploit the information from multi-camera networks for the monitoring of people and vehicles. Modules are proposed to perform critical surveillance tasks including: the management and calibration of cameras within a multi-camera network; tracking of objects across multiple views; recognition of people utilising biometrics and in particular soft-biometrics; the monitoring of crowds; and activity recognition. Recent advances in these computer vision modules and capability gaps in surveillance technology are also highlighted.

Class-specific sparse codes for representing activities

In this paper we investigate the effectiveness of class specific sparse codes in the context of discriminative action classification. The bag-of-words representation is widely used in activity recognition to encode features, and although it yields state-of-the art performance with several feature descriptors it still suffers from large quantization errors and reduces the overall performance. Recently proposed sparse representation methods have been shown to effectively represent features as a linear combination of an over complete dictionary by minimizing the reconstruction error. In contrast to most of the sparse representation methods which focus on Sparse-Reconstruction based Classification (SRC), this paper focuses on a discriminative classification using a SVM by constructing class-specific sparse codes for motion and appearance separately. Experimental results demonstrates that separate motion and appearance specific sparse coefficients provide the most effective and discriminative representation for each class compared to a single class-specific sparse coefficients.

Exchange relationships in building services maintenance activity : a test of transaction cost economics and the resource dependency view

The firm is faced with a decision concerning the nature of intra-organizational exchange relationships with internal human resources and the nature or inter-organizational exchange relationships with market firms. In both situations, the firm can develop an exchange that ranges from a discrete exchange to a relational exchange. Transaction Cost Economics (TCE) and the Resource Dependency View (RDV) represent alternative efficiency-based explanations fo the nature of the exchange relationship. The aim of the paper is to test these two theories in respect of air conditioning maintenance in retail centres. Multiple sources of information are genereated from case studies of Australian retail centres to test these theories in respoect of internalized operations management (concerning strategic aspects of air conditioning maintenance) and externalized planned routine air conditioning maintenance. The analysis of the data centres on pattern matching. It is concluded that the data supports TCE - on the basis of a development in TCE's contractual schema. Further research is suggested towards taking a pluralistic stance and developing a combined efficiency and power hypothesis - upon which Williamson has speculated. For practice, the conclusions also offer a timely cautionary note concerning the adoption of one approach in all exchange relationships.

Aligning policy with practice: an evaluation of vocational education and training in the library and information services sector

Vocational education and training for the library and information services (LIS) sector in Australia offers students the career pathway to become library technicians. Library technicians play a valuable role in drawing on sound practical knowledge and skills to support the delivery of library and information services that meet client needs. Over the past forty years, the Australian Library and Information Association (ALIA) has monitored the quality of library technician courses. Since 2005, ALIA has run national professional development days for library technician educators with the goal of establishing an alternative model for course recognition focusing on the process of peer review to benchmark good practice and stimulate continuous improvement in library technician education. This initial developmental work has culminated in 2009 with site visits to all library technician courses in Australia. The paper presents a whole-of-industry case study to critically review the work undertaken to date.

The delta-phase spectrum with application to voice activity detection and speaker recognition

For several reasons, the Fourier phase domain is less favored than the magnitude domain in signal processing and modeling of speech. To correctly analyze the phase, several factors must be considered and compensated, including the effect of the step size, windowing function and other processing parameters. Building on a review of these factors, this paper investigates a spectral representation based on the Instantaneous Frequency Deviation, but in which the step size between processing frames is used in calculating phase changes, rather than the traditional single sample interval. Reflecting these longer intervals, the term delta-phase spectrum is used to distinguish this from instantaneous derivatives. Experiments show that mel-frequency cepstral coefficients features derived from the delta-phase spectrum (termed Mel-Frequency delta-phase features) can produce broadly similar performance to equivalent magnitude domain features for both voice activity detection and speaker recognition tasks. Further, it is shown that the fusion of the magnitude and phase representations yields performance benefits over either in isolation.

Emergency health services : demand and service delivery models. Monograph 1: literature review and activity trends

Emergency Health Services (EHS), encompassing hospital-based Emergency Departments (ED) and pre-hospital ambulance services, are a significant and high profile component of Australia’s health care system and congestion of these, evidenced by physical overcrowding and prolonged waiting times, is causing considerable community and professional concern. This concern relates not only to Australia’s capacity to manage daily health emergencies but also the ability to respond to major incidents and disasters. EHS congestion is a result of the combined effects of increased demand for emergency care, increased complexity of acute health care, and blocked access to ongoing care (e.g. inpatient beds). Despite this conceptual understanding there is a lack of robust evidence to explain the factors driving increased demand, or how demand contributes to congestion, and therefore public policy responses have relied upon limited or unsound information. The Emergency Health Services Queensland (EHSQ) research program proposes to determine the factors influencing the growing demand for emergency health care and to establish options for alternative service provision that may safely meet patient’s needs. The EHSQ study is funded by the Australian Research Council (ARC) through its Linkage Program and is supported financially by the Queensland Ambulance Service (QAS). This monograph is part of a suite of publications based on the research findings that examines the existing literature, and current operational context. Literature was sourced using standard search approaches and a range of databases as well as a selection of articles cited in the reviewed literature. Public sources including the Australian Institute of Health and Welfare (AIHW), the Council of Ambulance Authorities (CAA) Annual Reports, Australian Bureau of Statistics (ABS) and Department of Health and Ageing (DoHA) were examined for trend data across Australia.