Information Exchange and Fusion in Dynamic and Heterogeneous Distributed Environments

Context awareness, dynamic reconfiguration at runtime and heterogeneity are key characteristics of future distributed systems, particularly in ubiquitous and mobile computing scenarios. The main contributions of this dissertation are theoretical as well as architectural concepts facilitating information exchange and fusion in heterogeneous and dynamic distributed environments. Our main focus is on bridging the heterogeneity issues and, at the same time, considering uncertain, imprecise and unreliable sensor information in information fusion and reasoning approaches. A domain ontology is used to establish a common vocabulary for the exchanged information. We thereby explicitly support different representations for the same kind of information and provide Inter-Representation Operations that convert between them. Special account is taken of the conversion of associated meta-data that express uncertainty and impreciseness. The Unscented Transformation, for example, is applied to propagate Gaussian normal distributions across highly non-linear Inter-Representation Operations. Uncertain sensor information is fused using the Dempster-Shafer Theory of Evidence as it allows explicit modelling of partial and complete ignorance. We also show how to incorporate the Dempster-Shafer Theory of Evidence into probabilistic reasoning schemes such as Hidden Markov Models in order to be able to consider the uncertainty of sensor information when deriving high-level information from low-level data. For all these concepts we provide architectural support as a guideline for developers of innovative information exchange and fusion infrastructures that are particularly targeted at heterogeneous dynamic environments. Two case studies serve as proof of concept. The first case study focuses on heterogeneous autonomous robots that have to spontaneously form a cooperative team in order to achieve a common goal. The second case study is concerned with an approach for user activity recognition which serves as baseline for a context-aware adaptive application. Both case studies demonstrate the viability and strengths of the proposed solution and emphasize that the Dempster-Shafer Theory of Evidence should be preferred to pure probability theory in applications involving non-linear Inter-Representation Operations.

Learning and detecting activities from movement trajectories using the hierarchical hidden Markov model

Directly modeling the inherent hierarchy and shared structures of human behaviors, we present an application of the hierarchical hidden Markov model (HHMM) for the problem of activity recognition. We argue that to robustly model and recognize complex human activities, it is crucial to exploit both the natural hierarchical decomposition and shared semantics embedded in the movement trajectories. To this end, we propose the use of the HHMM, a rich stochastic model that has been recently extended to handle shared structures, for representing and recognizing a set of complex indoor activities. Furthermore, in the need of real-time recognition, we propose a Rao-Blackwellised particle filter (RBPF) that efficiently computes the filtering distribution at a constant time complexity for each new observation arrival. The main contributions of this paper lie in the application of the shared-structure HHMM, the estimation of the model's parameters at all levels simultaneously, and a construction of an RBPF approximate inference scheme. The experimental results in a real-world environment have confirmed our belief that directly modeling shared structures not only reduces computational cost, but also improves recognition accuracy when compared with the tree HHMM and the flat HMM.

Utility-based summarization of home videos

The aim of this work is to devise an effective method for static summarization of home video sequences. Based on the premise that the user watching a summary is interested in people related (how many, who, emotional state) or activity related aspects, we formulate a novel approach to video summarization that works to specifically expose relevant video frames that make the content spotting tasks possible. Unlike existing approaches, which work on low-level features which often produce the summary not appealing to the viewer due to the semantic gap between low-level features and high-level concepts, our approach is driven by various utility functions (identity count, identity recognition, emotion recognition, activity recognition, sense of space) that use the results of face detection, face clustering, shot clustering and within cluster frame alignment. The summarization problem is then treated as the problem of extracting the set of key frames that have the maximum combined utility.

Learning latent activities from social signals with hierarchical dirichlet processes

Understanding human activities is an important research topic, most noticeably in assisted-living and healthcare monitoring environments. Beyond simple forms of activity (e.g., an RFID event of entering a building), learning latent activities that are more semantically interpretable, such as sitting at a desk, meeting with people, or gathering with friends, remains a challenging problem. Supervised learning has been the typical modeling choice in the past. However, this requires labeled training data, is unable to predict never-seen-before activity, and fails to adapt to the continuing growth of data over time. In this chapter, we explore the use of a Bayesian nonparametric method, in particular the hierarchical Dirichlet process, to infer latent activities from sensor data acquired in a pervasive setting. Our framework is unsupervised, requires no labeled data, and is able to discover new activities as data grows. We present experiments on extracting movement and interaction activities from sociometric badge signals and show how to use them for detecting of subcommunities. Using the popular Reality Mining dataset, we further demonstrate the extraction of colocation activities and use them to automatically infer the structure of social subgroups. © 2014 Elsevier Inc. All rights reserved.

Nonparametric discovery of movement patterns from accelerometer signals

Monitoring daily physical activity plays an important role in disease prevention and intervention. This paper proposes an approach to monitor the body movement intensity levels from accelerometer data. We collect the data using the accelerometer in a realistic setting without any supervision. The ground-truth of activities is provided by the participants themselves using an experience sampling application running on their mobile phones. We compute a novel feature that has a strong correlation with the movement intensity. We use the hierarchical Dirichlet process (HDP) model to detect the activity levels from this feature. Consisting of Bayesian nonparametric priors over the parameters the model can infer the number of levels automatically. By demonstrating the approach on the publicly available USC-HAD dataset that includes ground-truth activity labels, we show a strong correlation between the discovered activity levels and the movement intensity of the activities. This correlation is further confirmed using our newly collected dataset. We further use the extracted patterns as features for clustering and classifying the activity sequences to improve performance.

Context detection and abstraction in smart environments

Context-aware computing is currently considered the most promising approach to overcome information overload and to speed up access to relevant information and services. Context-awareness may be derived from many sources, including user profile and preferences, network information, sensor analysis; usually context-awareness relies on the ability of computing devices to interact with the physical world, i.e. with the natural and artificial objects hosted within the "environment”. Ideally, context-aware applications should not be intrusive and should be able to react according to user’s context, with minimum user effort. Context is an application dependent multidimensional space and the location is an important part of it since the very beginning. Location can be used to guide applications, in providing information or functions that are most appropriate for a specific position. Hence location systems play a crucial role. There are several technologies and systems for computing location to a vary degree of accuracy and tailored for specific space model, i.e. indoors or outdoors, structured spaces or unstructured spaces. The research challenge faced by this thesis is related to pedestrian positioning in heterogeneous environments. Particularly, the focus will be on pedestrian identification, localization, orientation and activity recognition. This research was mainly carried out within the “mobile and ambient systems” workgroup of EPOCH, a 6FP NoE on the application of ICT to Cultural Heritage. Therefore applications in Cultural Heritage sites were the main target of the context-aware services discussed. Cultural Heritage sites are considered significant test-beds in Context-aware computing for many reasons. For example building a smart environment in museums or in protected sites is a challenging task, because localization and tracking are usually based on technologies that are difficult to hide or harmonize within the environment. Therefore it is expected that the experience made with this research may be useful also in domains other than Cultural Heritage. This work presents three different approaches to the pedestrian identification, positioning and tracking: Pedestrian navigation by means of a wearable inertial sensing platform assisted by the vision based tracking system for initial settings an real-time calibration; Pedestrian navigation by means of a wearable inertial sensing platform augmented with GPS measurements; Pedestrian identification and tracking, combining the vision based tracking system with WiFi localization. The proposed localization systems have been mainly used to enhance Cultural Heritage applications in providing information and services depending on the user’s actual context, in particular depending on the user’s location.

Progetto di reti Sensori Wireless e tecniche di Fusione Sensoriale

Ambient Intelligence (AmI) envisions a world where smart, electronic environments are aware and responsive to their context. People moving into these settings engage many computational devices and systems simultaneously even if they are not aware of their presence. AmI stems from the convergence of three key technologies: ubiquitous computing, ubiquitous communication and natural interfaces. The dependence on a large amount of fixed and mobile sensors embedded into the environment makes of Wireless Sensor Networks one of the most relevant enabling technologies for AmI. WSN are complex systems made up of a number of sensor nodes, simple devices that typically embed a low power computational unit (microcontrollers, FPGAs etc.), a wireless communication unit, one or more sensors and a some form of energy supply (either batteries or energy scavenger modules). Low-cost, low-computational power, low energy consumption and small size are characteristics that must be taken into consideration when designing and dealing with WSNs. In order to handle the large amount of data generated by a WSN several multi sensor data fusion techniques have been developed. The aim of multisensor data fusion is to combine data to achieve better accuracy and inferences than could be achieved by the use of a single sensor alone. In this dissertation we present our results in building several AmI applications suitable for a WSN implementation. The work can be divided into two main areas: Multimodal Surveillance and Activity Recognition. Novel techniques to handle data from a network of low-cost, low-power Pyroelectric InfraRed (PIR) sensors are presented. Such techniques allow the detection of the number of people moving in the environment, their direction of movement and their position. We discuss how a mesh of PIR sensors can be integrated with a video surveillance system to increase its performance in people tracking. Furthermore we embed a PIR sensor within the design of a Wireless Video Sensor Node (WVSN) to extend its lifetime. Activity recognition is a fundamental block in natural interfaces. A challenging objective is to design an activity recognition system that is able to exploit a redundant but unreliable WSN. We present our activity in building a novel activity recognition architecture for such a dynamic system. The architecture has a hierarchical structure where simple nodes performs gesture classification and a high level meta classifiers fuses a changing number of classifier outputs. We demonstrate the benefit of such architecture in terms of increased recognition performance, and fault and noise robustness. Furthermore we show how we can extend network lifetime by performing a performance-power trade-off. Smart objects can enhance user experience within smart environments. We present our work in extending the capabilities of the Smart Micrel Cube (SMCube), a smart object used as tangible interface within a tangible computing framework, through the development of a gesture recognition algorithm suitable for this limited computational power device. Finally the development of activity recognition techniques can greatly benefit from the availability of shared dataset. We report our experience in building a dataset for activity recognition. Such dataset is freely available to the scientific community for research purposes and can be used as a testbench for developing, testing and comparing different activity recognition techniques.

Riconoscimento automatico di attività attraverso i Google Play Services: una valutazione sperimentale

Tesi sperimentale per testare la nuova libreria di Activity Recognition per Android di Google inc.

Progettazione ed implementazione di un'applicazione Android basata sul riconoscimento automatico di attività

Progettazione ed implementazione di una semplice applicazione per smartphone Android al fine di dimostrare le funzionalità delle librerie per l'activity recognition messe a disposizione dai Google Play Services. Lo studio esplora il campo di ricerca in generale, mostrandone le modalità, le applicazioni e le problematiche, e introduce l'ambiente Android per poi analizzare l'applicazione progettata. In conclusione, vengono mostrati alcuni test svolti per verificare l'accuratezza del classificatore implementato da Google.

Progettazione di un applicativo di supporto al monitoraggio di team situati

Lo scopo della tesi e quello di giungere alla realizzazione di un prototipo, che si basi su tecnologie specifiche quali GPS, Android e Bluetooth, per l'infrastruttura di un sistema che può essere visto come l'unione di tre macro parti distinte, centrale di controllo, dispositivo mobile e pulsossimetro. Concentrandosi in particolare sugli ultimi due componenti citati e realizzando un sistema di comunicazione che si basa su un Web Service ispirato al modello REST, si giungerà, attraverso un attento processo logico dettato dai canoni dell'ingegneria del software, al prototipo finale. Il prototipo che sarà realizzato rappresenterà oltre che un primo sistema funzionante e operativo, un punto di partenza per estensioni future, con lo scopo di perfezionare le funzionalità già esistenti o di fornirne di aggiuntive.

Towards Proactive Context-aware Computing and Systems

A primary goal of context-aware systems is delivering the right information at the right place and right time to users in order to enable them to make effective decisions and improve their quality of life. There are three key requirements for achieving this goal: determining what information is relevant, personalizing it based on the users’ context (location, preferences, behavioral history etc.), and delivering it to them in a timely manner without an explicit request from them. These requirements create a paradigm that we term as “Proactive Context-aware Computing”. Most of the existing context-aware systems fulfill only a subset of these requirements. Many of these systems focus only on personalization of the requested information based on users’ current context. Moreover, they are often designed for specific domains. In addition, most of the existing systems are reactive - the users request for some information and the system delivers it to them. These systems are not proactive i.e. they cannot anticipate users’ intent and behavior and act proactively without an explicit request from them. In order to overcome these limitations, we need to conduct a deeper analysis and enhance our understanding of context-aware systems that are generic, universal, proactive and applicable to a wide variety of domains. To support this dissertation, we explore several directions. Clearly the most significant sources of information about users today are smartphones. A large amount of users’ context can be acquired through them and they can be used as an effective means to deliver information to users. In addition, social media such as Facebook, Flickr and Foursquare provide a rich and powerful platform to mine users’ interests, preferences and behavioral history. We employ the ubiquity of smartphones and the wealth of information available from social media to address the challenge of building proactive context-aware systems. We have implemented and evaluated a few approaches, including some as part of the Rover framework, to achieve the paradigm of Proactive Context-aware Computing. Rover is a context-aware research platform which has been evolving for the last 6 years. Since location is one of the most important context for users, we have developed ‘Locus’, an indoor localization, tracking and navigation system for multi-story buildings. Other important dimensions of users’ context include the activities that they are engaged in. To this end, we have developed ‘SenseMe’, a system that leverages the smartphone and its multiple sensors in order to perform multidimensional context and activity recognition for users. As part of the ‘SenseMe’ project, we also conducted an exploratory study of privacy, trust, risks and other concerns of users with smart phone based personal sensing systems and applications. To determine what information would be relevant to users’ situations, we have developed ‘TellMe’ - a system that employs a new, flexible and scalable approach based on Natural Language Processing techniques to perform bootstrapped discovery and ranking of relevant information in context-aware systems. In order to personalize the relevant information, we have also developed an algorithm and system for mining a broad range of users’ preferences from their social network profiles and activities. For recommending new information to the users based on their past behavior and context history (such as visited locations, activities and time), we have developed a recommender system and approach for performing multi-dimensional collaborative recommendations using tensor factorization. For timely delivery of personalized and relevant information, it is essential to anticipate and predict users’ behavior. To this end, we have developed a unified infrastructure, within the Rover framework, and implemented several novel approaches and algorithms that employ various contextual features and state of the art machine learning techniques for building diverse behavioral models of users. Examples of generated models include classifying users’ semantic places and mobility states, predicting their availability for accepting calls on smartphones and inferring their device charging behavior. Finally, to enable proactivity in context-aware systems, we have also developed a planning framework based on HTN planning. Together, these works provide a major push in the direction of proactive context-aware computing.

Modelling human behaviour and event detection from videos

Automatic analysis of human behaviour in large collections of videos is gaining interest, even more so with the advent of file sharing sites such as YouTube. However, challenges still exist owing to several factors such as inter- and intra-class variations, cluttered backgrounds, occlusion, camera motion, scale, view and illumination changes. This research focuses on modelling human behaviour for action recognition in videos. The developed techniques are validated on large scale benchmark datasets and applied on real-world scenarios such as soccer videos. Three major contributions are made. The first contribution is in the area of proper choice of a feature representation for videos. This involved a study of state-of-the-art techniques for action recognition, feature extraction processing and dimensional reduction techniques so as to yield the best performance with optimal computational requirements. Secondly, temporal modelling of human behaviour is performed. This involved frequency analysis and temporal integration of local information in the video frames to yield a temporal feature vector. Current practices mostly average the frame information over an entire video and neglect the temporal order. Lastly, the proposed framework is applied and further adapted to real-world scenario such as soccer videos. A dataset consisting of video sequences depicting events of players falling is created from actual match data to this end and used to experimentally evaluate the proposed framework.

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.