PhD forum : multiple camera management using wide base-line matching

Manual calibration of large and dynamic networks of cameras is labour intensive and time consuming. This is a strong motivator for the development of automatic calibration methods. Automatic calibration relies on the ability to find correspondences between multiple views of the same scene. If the cameras are sparsely placed, this can be a very difficult task. This PhD project focuses on the further development of uncalibrated wide baseline matching techniques.

Listening to nature : techniques for large-scale monitoring of ecosystems using acoustics

Climate change and human activity are subjecting the environment to unprecedented rates of change. Monitoring these changes is an immense task that demands new levels of automated monitoring and analysis. We propose the use of acoustics as a proxy for the time consuming auditing of fauna, especially for determining the presence/absence of species. Acoustic monitoring is deceptively simple; seemingly all that is required is a sound recorder. However there are many major challenges if acoustics are to be used for large scale monitoring of ecosystems. Key issues are scalability and automation. This paper discusses our approach to this important research problem. Our work is being undertaken in collaboration with ecologists interested both in identifying particular species and in general ecosystem health.

Listening to nature : acoustic monitoring of the environment

The requirement to monitor the rapid pace of environmental change due to global warming and to human development is producing large volumes of data but placing much stress on the capacity of ecologists to store, analyse and visualise that data. To date, much of the data has been provided by low level sensors monitoring soil moisture, dissolved nutrients, light intensity, gas composition and the like. However, a significant part of an ecologist’s work is to obtain information about species diversity, distributions and relationships. This task typically requires the physical presence of an ecologist in the field, listening and watching for species of interest. It is an extremely difficult task to automate because of the higher order difficulties in bandwidth, data management and intelligent analysis if one wishes to emulate the highly trained eyes and ears of an ecologist. This paper is concerned with just one part of the bigger challenge of environmental monitoring – the acquisition and analysis of acoustic recordings of the environment. Our intention is to provide helpful tools to ecologists – tools that apply information technologies and computational technologies to all aspects of the acoustic environment. The on-line system which we are building in conjunction with ecologists offers an integrated approach to recording, data management and analysis. The ecologists we work with have different requirements and therefore we have adopted the toolbox approach, that is, we offer a number of different web services that can be concatenated according to need. In particular, one group of ecologists is concerned with identifying the presence or absence of species and their distributions in time and space. Another group, motivated by legislative requirements for measuring habitat condition, are interested in summary indices of environmental health. In both case, the key issues are scalability and automation.

Improved maritime target tracker using colour fusion

Searching for humans lost in vast stretches of ocean has always been a difficult task. This paper investigates a machine vision system that addresses this problem by exploiting the useful properties of alternate colour spaces. In particular, the paper investigates the fusion of colour information from the HSV, RGB, YCbCr and YIQ colour spaces within the emission matrix of a Hidden Markov Model tracker to enhance video based maritime target detection. The system has shown promising results. The paper also identifies challenges still needing to be met.

Comparison of 3D finite element analysis derived stiffness and BMD to determine the failure load of the excised proximal femur

Introduction: Bone mineral density (BMD) is currently the preferred surrogate for bone strength in clinical practice. Finite element analysis (FEA) is a computer simulation technique that can predict the deformation of a structure when a load is applied, providing a measure of stiffness (Nmm−1). Finite element analysis of X-ray images (3D-FEXI) is a FEA technique whose analysis is derived froma single 2D radiographic image. Methods: 18 excised human femora had previously been quantitative computed tomography scanned, from which 2D BMD-equivalent radiographic images were derived, and mechanically tested to failure in a stance-loading configuration. A 3D proximal femur shape was generated from each 2D radiographic image and used to construct 3D-FEA models. Results: The coefficient of determination (R2%) to predict failure load was 54.5% for BMD and 80.4% for 3D-FEXI. Conclusions: This ex vivo study demonstrates that 3D-FEXI derived from a conventional 2D radiographic image has the potential to significantly increase the accuracy of failure load assessment of the proximal femur compared with that currently achieved with BMD. This approach may be readily extended to routine clinical BMD images derived by dual energy X-ray absorptiometry. Crown Copyright © 2009 Published by Elsevier Ltd on behalf of IPEM. All rights reserved

Relative entropy rate based multiple hidden Markov Model Approximation

This paper proposes a novel relative entropy rate (RER) based approach for multiple HMM (MHMM) approximation of a class of discrete-time uncertain processes. Under different uncertainty assumptions, the model design problem is posed either as a min-max optimisation problem or stochastic minimisation problem on the RER between joint laws describing the state and output processes (rather than the more usual RER between output processes). A suitable filter is proposed for which performance results are established which bound conditional mean estimation performance and show that estimation performance improves as the RER is reduced. These filter consistency and convergence bounds are the first results characterising multiple HMM approximation performance and suggest that joint RER concepts provide a useful model selection criteria. The proposed model design process and MHMM filter are demonstrated on an important image processing dim-target detection problem.

A genetic algorithm for the multi-source and multi-sink minimum vertex cut problem and its applications

We present a new penalty-based genetic algorithm for the multi-source and multi-sink minimum vertex cut problem, and illustrate the algorithm’s usefulness with two real-world applications. It is proved in this paper that the genetic algorithm always produces a feasible solution by exploiting some domain-specific knowledge. The genetic algorithm has been implemented on the example applications and evaluated to show how well it scales as the problem size increases.

High-level concept annotation using ontology and probabilistic inference

Image annotation is a significant step towards semantic based image retrieval. Ontology is a popular approach for semantic representation and has been intensively studied for multimedia analysis. However, relations among concepts are seldom used to extract higher-level semantics. Moreover, the ontology inference is often crisp. This paper aims to enable sophisticated semantic querying of images, and thus contributes to 1) an ontology framework to contain both visual and contextual knowledge, and 2) a probabilistic inference approach to reason the high-level concepts based on different sources of information. The experiment on a natural scene database from LabelMe database shows encouraging results.

Prognosis of bearing failure based on health state estimation

This paper proposes a new prognosis model based on the technique for health state estimation of machines for accurate assessment of the remnant life. For the evaluation of health stages of machines, the Support Vector Machine (SVM) classifier was employed to obtain the probability of each health state. Two case studies involving bearing failures were used to validate the proposed model. Simulated bearing failure data and experimental data from an accelerated bearing test rig were used to train and test the model. The result obtained is very encouraging and shows that the proposed prognostic model produces promising results and has the potential to be used as an estimation tool for machine remnant life prediction.

Towards executable specification : combining i* and AgentSpeak(L).

Agent-oriented conceptual modelling (AoCM) approaches in Requirements Engineering (RE) have received considerable attention recently. Semi-formal modeling frameworks such as i* assist analysts in requirements elicitation and reasoning of early-phase RE. AgentSpeak(L) is a widely accepted agent programming language. The Strategic Rationale (SR) model of the i* framework naturally lends itself to AgentSpeak(L) programs. Furthermore, the Strategic Dependency (SD) component of the i* framework prescribes the interaction between the agents in a multi-agent environment. This paper proposes a formal methodology for transforming a SR model to an AgentS- peak(L) agent. The constructed AgentSpeak(L) agents will then form the essential components of a multi-agent system, MAS.

Adaptive unsupervised learning of human actions

Automatic detection of suspicious activities in CCTV camera feeds is crucial to the success of video surveillance systems. Such a capability can help transform the dumb CCTV cameras into smart surveillance tools for fighting crime and terror. Learning and classification of basic human actions is a precursor to detecting suspicious activities. Most of the current approaches rely on a non-realistic assumption that a complete dataset of normal human actions is available. This paper presents a different approach to deal with the problem of understanding human actions in video when no prior information is available. This is achieved by working with an incomplete dataset of basic actions which are continuously updated. Initially, all video segments are represented by Bags-Of-Words (BOW) method using only Term Frequency-Inverse Document Frequency (TF-IDF) features. Then, a data-stream clustering algorithm is applied for updating the system's knowledge from the incoming video feeds. Finally, all the actions are classified into different sets. Experiments and comparisons are conducted on the well known Weizmann and KTH datasets to show the efficacy of the proposed approach.

Crowd Counting Using Multiple Local Features

In public venues, crowd size is a key indicator of crowd safety and stability. Crowding levels can be detected using holistic image features, however this requires a large amount of training data to capture the wide variations in crowd distribution. If a crowd counting algorithm is to be deployed across a large number of cameras, such a large and burdensome training requirement is far from ideal. In this paper we propose an approach that uses local features to count the number of people in each foreground blob segment, so that the total crowd estimate is the sum of the group sizes. This results in an approach that is scalable to crowd volumes not seen in the training data, and can be trained on a very small data set. As a local approach is used, the proposed algorithm can easily be used to estimate crowd density throughout different regions of the scene and be used in a multi-camera environment. A unique localised approach to ground truth annotation reduces the required training data is also presented, as a localised approach to crowd counting has different training requirements to a holistic one. Testing on a large pedestrian database compares the proposed technique to existing holistic techniques and demonstrates improved accuracy, and superior performance when test conditions are unseen in the training set, or a minimal training set is used.

Selecting, optimizing and fusing ‘salient’ Gabor features for facial expression recognition

This paper describes a novel framework for facial expression recognition from still images by selecting, optimizing and fusing ‘salient’ Gabor feature layers to recognize six universal facial expressions using the K nearest neighbor classifier. The recognition comparisons with all layer approach using JAFFE and Cohn-Kanade (CK) databases confirm that using ‘salient’ Gabor feature layers with optimized sizes can achieve better recognition performance and dramatically reduce computational time. Moreover, comparisons with the state of the art performances demonstrate the effectiveness of our approach.

QoS-Based Web Service Scheduling Accommodating Inter-Service Dependencies Using Minimal-Conflict Hill-Climbing Repair Genetic Algorithm

In the filed of semantic grid, QoS-based Web service scheduling for workflow optimization is an important problem.However, in semantic and service rich environment like semantic grid, the emergence of context constraints on Web services is very common making the scheduling consider not only quality properties of Web services, but also inter service dependencies which are formed due to the context constraints imposed on Web services. In this paper, we present a repair genetic algorithm, namely minimal-conflict hill-climbing repair genetic algorithm, to address scheduling optimization problems in workflow applications in the presence of domain constraints and inter service dependencies. Experimental results demonstrate the scalability and effectiveness of the genetic algorithm.

A Penalty-based Genetic Algorithm for QoS-AwareWeb Service Composition with Inter-Service Dependencies and Conflicts

In Web service based systems, new value-added Web services can be constructed by integrating existing Web services. A Web service may have many implementations, which are functionally identical, but have different Quality of Service (QoS) attributes, such as response time, price, reputation, reliability, availability and so on. Thus, a significant research problem in Web service composition is how to select an implementation for each of the component Web services so that the overall QoS of the composite Web service is optimal. This is so called QoS-aware Web service composition problem. In some composite Web services there are some dependencies and conflicts between the Web service implementations. However, existing approaches cannot handle the constraints. This paper tackles the QoS-aware Web service composition problem with inter service dependencies and conflicts using a penalty-based genetic algorithm (GA). Experimental results demonstrate the effectiveness and the scalability of the penalty-based GA.