Inter-Frame Contextual Modelling For Visual Speech Recognition

In this paper, we present a new approach to visual speech recognition which improves contextual modelling by combining Inter-Frame Dependent and Hidden Markov Models. This approach captures contextual information in visual speech that may be lost using a Hidden Markov Model alone. We apply contextual modelling to a large speaker independent isolated digit recognition task, and compare our approach to two commonly adopted feature based techniques for incorporating speech dynamics. Results are presented from baseline feature based systems and the combined modelling technique. We illustrate that both of these techniques achieve similar levels of performance when used independently. However significant improvements in performance can be achieved through a combination of the two. In particular we report an improvement in excess of 17% relative Word Error Rate in comparison to our best baseline system.

The biology and ecology of the ocean sunfish Mola mola: a review of current knowledge and future research perspectives

Relatively little is known about the biology and ecology of the world's largest (heaviest) bony fish, the ocean sunfish Mola mola, despite its worldwide occurrence in temperate and tropical seas. Studies are now emerging that require many common perceptions about sunfish behaviour and ecology to be re-examined. Indeed, the long-held view that ocean sunfish are an inactive, passively drifting species seems to be entirely misplaced. Technological advances in marine telemetry are revealing distinct behavioural patterns and protracted seasonal movements. Extensive forays by ocean sunfish into the deep ocean have been documented and broad-scale surveys, together with molecular and laboratory based techniques, are addressing the connectivity and trophic role of these animals. These emerging molecular and movement studies suggest that local distinct populations may be prone to depletion through bycatch in commercial fisheries. Rising interest in ocean sunfish, highlighted by the increase in recent publications, warrants a thorough review of the biology and ecology of this species. Here we review the taxonomy, morphology, geography, diet, locomotion, vision, movements, foraging ecology, reproduction and species interactions of M. mola. We present a summary of current conservation issues and suggest methods for addressing fundamental gaps in our knowledge.

Developing a simple, rapid method for identifying and monitoring jellyfish aggregations from the air

Within the marine environment, aerial surveys have historically centred on apex predators, such as pinnipeds, cetaceans and sea birds. However, it is becoming increasingly apparent that the utility of this technique may also extend to subsurface species such as pre-spawning fish stocks and aggregations of jellyfish that occur close to the surface. In light of this, we tested the utility of aerial surveys to provide baseline data for 3 poorly understood scyphozoan jellyfish found throughout British and Irish waters: Rhizostoma octopus, Cyanea capillata and Chrysaora hysoscella. Our principal objectives were to develop a simple sampling protocol to identify and quantify surface aggregations, assess their consistency in space and time, and consider the overall applicability of this technique to the study of gelatinous zooplankton. This approach provided a general understanding of range and relative abundance for each target species, with greatest suitability to the study of R. octopus. For this species it was possible to identify and monitor extensive, temporally consistent and previously undocumented aggregations throughout the Irish Sea, an area spanning thousands of square kilometres. This finding has pronounced implications for ecologists and fisheries managers alike and, moreover, draws attention to the broad utility of aerial surveys for the study of gelatinous aggregations beyond the range of conventional ship-based techniques.

A structural analysis of G-quadruplex/ligand interactions

This focused review article discusses in detail, all available high-resolution small molecule ligand/G-quadruplex structural data derived from crystallographic and NMR based techniques, in an attempt to understand key factors in ligand binding and to highlight the biological importance of these complexes. In contrast to duplex DNA, G-quadruplexes are four-stranded nucleic acid structures folded from guanine rich repeat sequences stabilized by the stacking of guanine G-quartets and extensive Watson-Crick/Hoogsteen hydrogen bonding. Thermally stable, these topologies can play a role in telomere regulation and gene expression. The core structures of G-quadruplexes form stable scaffolds while the loops have been shown, by the addition of small molecule ligands, to be sufficiently adaptable to generate new and extended binding platforms for ligands to associate, either by extending G-quartet surfaces or by forming additional planar dinucleotide pairings. Many of these structurally characterised loop rearrangements were totally unexpected opening up new opportunities for the design of selective ligands. However these rearrangements do significantly complicate attempts to rationally design ligands against well defined but unbound topologies, as seen for the series of napthalene diimides complexes. Drawing together previous findings and with the introduction of two new crystallographic quadruplex/ligand structures we aim to expand the understanding of possible structural adaptations available to quadruplexes in the presence of ligands, thereby aiding in the design of new selective entities. (C) 2011 Elsevier Masson SAS. All rights reserved.

Multicenter Blinded Analysis of RT-PCR Detection Methods for Paramyxoviruses in Relation to Paget's Disease of Bone

Conflicting results have been reported on the detection of paramyxovirus transcripts in Paget's disease, and a possible explanation is differences in the sensitivity of RT-PCR methods for detecting virus. In a blinded study, we found no evidence to suggest that laboratories that failed to detect viral transcripts had less sensitive RT-PCR assays, and we did not detect measles or distemper transcripts in Paget's samples using the most sensitive assays evaluated.

Introduction: There is conflicting evidence on the possible role of persistent paramyxovirus infection in Paget's disease of bone (PDB). Some workers have detected measles virus (MV) or canine distemper virus (CDV) transcripts in cells and tissues from patients with PDB, but others have failed to confirm this finding. A possible explanation might be differences in the sensitivity of RT-PCR methods for detecting virus. Here we performed a blinded comparison of the sensitivity of different RT-PCR-based techniques for MV and CDV detection in different laboratories and used the most sensitive assays to screen for evidence of viral transcripts in bone and blood samples derived from patients with PDB.

Materials and Methods: Participating laboratories analyzed samples spiked with known amounts of MV and CDV transcripts and control samples that did not contain viral nucleic acids. All analyses were performed on a blinded basis.

Results: The limit of detection for CDV was 1000 viral transcripts in three laboratories (Aberdeen, Belfast, and Liverpool) and 10,000 transcripts in another laboratory (Manchester). The limit of detection for MV was 16 transcripts in one laboratory (NIBSC), 1000 transcripts in two laboratories (Aberdeen and Belfast), and 10,000 transcripts in two laboratories (Liverpool and Manchester). An assay previously used by a U.S.-based group to detect MV transcripts in PDB had a sensitivity of 1000 transcripts. One laboratory (Manchester) detected CDV transcripts in a negative control and in two samples that had been spiked with MV. None of the other laboratories had false-positive results for MV or CDV, and no evidence of viral transcripts was found on analysis of 12 PDB samples using the most sensitive RT-PCR assays for MV and CDV.

Conclusions: We found that RT-PCR assays used by different laboratories differed in their sensitivity to detect CDV and MV transcripts but found no evidence to suggest that laboratories that previously failed to detect viral transcripts had less sensitive RT-PCR assays than those that detected viral transcripts. False-positive results were observed with one laboratory, and we failed to detect paramyxovirus transcripts in PDB samples using the most sensitive assays evaluated. Our results show that failure of some laboratories to detect viral transcripts is unlikely to be caused by problems with assay sensitivity and highlight the fact that contamination can be an issue when searching for pathogens by sensitive RT-PCR-based techniques.

Starbog: The Garden Virtuoso and the Turing Machine

This article will discuss a recent ensemble composition entitled Starbog which was toured and broadcast in Britain in 2006 . The composition of Starbog focused on developing working methods which combined computer-based techniques (using OpenMusic) with more subconscious means of generating musical ideas. The challenge in achieving this was as much aesthetic/philosophical as it was technical and the present article is intending as a ‘sounding’ which focuses on the influence OpenMusic has had on the composer’s music, rather than documenting the nature of the often simple application of algorithms.

Complex Interactions between Dioxin-Like and Non-Dioxin-Like Compounds for in Vitro Cellular Responses: Implications for the Identification of Dioxin Exposure Biomarkers

Despite considerable advances in reducing the production of dioxin-like toxicants in recent years, contamination of the food chain still occasionally occurs resulting in huge losses to the agri-food sector and risk to human health through exposure. Dioxin-like toxicity is exhibited by a range of stable and bioaccumulative compounds including polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs), produced by certain types of combustion, and man-made coplanar polychlorinated biphenyls (PCBs), as found in electrical transformer oils. While dioxinergic compounds act by a common mode of action making exposure detection biomarker based techniques a potentially useful tool, the influence of co-contaminating toxicants on such approaches needs to be considered. To assess the impact of possible interactions, the biological responses of H4IIE cells to challenge by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in combination with PCB-52 and benzo-a-pyrene (BaP) were evaluated by a number of methods in this study. Ethoxyresorufin-O-deethylase (EROD) induction in TCDD exposed cells was suppressed by increasing concentrations of PCB-52, PCB-153, or BaP up to 10 mu M. BaP levels below 1 mu M suppressed TCDD stimulated EROD induction, but at higher concentrations, EROD induction was greater than the maximum observed when cells were treated with TCDD alone. A similar biphasic interaction of BaP with TCDD co-exposure was noted in the AlamarBlue assay and to a lesser extent with PCB-52. Surface enhanced laser desorption/ionization-time of flight mass spectrometry (SELDI-TOF) profiling of peptidomic responses of cells exposed to compound combinations was compared. Cells co-exposed to TCDD in the presence of BaP or PCB-52 produced the most differentiated spectra with a substantial number of non-additive interactions observed. These findings suggest that interactions between dioxin and other toxicants create novel, additive, and non-additive effects, which may be more indicative of the types of responses seen in exposed animals than those of single exposures to the individual compounds.

The formal specification of abstract data types and their implementation in Fortran 90: Implementation issues concerning the use of pointers

In this paper we continue our investigation into the development of computational-science software based on the identification and formal specification of Abstract Data Types (ADTs) and their implementation in Fortran 90. In particular, we consider the consequences of using pointers when implementing a formally specified ADT in Fortran 90. Our aim is to highlight the resulting conflict between the goal of information hiding, which is central to the ADT methodology, and the space efficiency of the implementation. We show that the issue of storage recovery cannot be avoided by the ADT user, and present a range of implementations of a simple ADT to illustrate various approaches towards satisfactory storage management. Finally, we propose a set of guidelines for implementing ADTs using pointers in Fortran 90. These guidelines offer a way gracefully to provide disposal operations in Fortran 90. Such an approach is desirable since Fortran 90 does not provide automatic garbage collection which is offered by many object-oriented languages including Eiffel, Java, Smalltalk, and Simula.

Evolving to the optoelectronic mouse for phycotoxin analysis in shellfish

Despite ethical and technical concerns, the in vivo method, or more commonly referred to mouse bioassay (MBA), is employed globally as a reference method for phycotoxin analysis in shellfish. This is particularly the case for paralytic shellfish poisoning (PSP) and emerging toxin monitoring. A high-performance liquid chromatography method (HPLC-FLD) has been developed for PSP toxin analysis, but due to difficulties and limitations in the method, this procedure has not been fully implemented as a replacement. Detection of the diarrhetic shellfish poisoning (DSP) toxins has moved towards LC-mass spectrometry (MS) analysis, whereas the analysis of the amnesic shellfish poisoning (ASP) toxin domoic acid is performed by HPLC. Although alternative methods of detection to the MBA have been described, each procedure is specific for a particular toxin and its analogues, with each group of toxins requiring separate analysis utilising different extraction procedures and analytical equipment. In addition, consideration towards the detection of unregulated and emerging toxins on the replacement of the MBA must be given. The ideal scenario for the monitoring of phycotoxins in shellfish and seafood would be to evolve to multiple toxin detection on a single bioanalytical sensing platform, i.e. 'an artificial mouse'. Immunologically based techniques and in particular surface plasmon resonance technology have been shown as a highly promising bioanalytical tool offering rapid, real-time detection requiring minimal quantities of toxin standards. A Biacore Q and a prototype multiplex SPR biosensor have been evaluated for their ability to be fit for purpose for the simultaneous detection of key regulated phycotoxin groups and the emerging toxin palytoxin. Deemed more applicable due to the separate flow channels, the prototype performance for domoic acid, okadaic acid, saxitoxin, and palytoxin calibration curves in shellfish achieved detection limits (IC20) of 4,000, 36, 144 and 46 μg/kg of mussel, respectively. A one-step extraction procedure demonstrated recoveries greater than 80 % for all toxins. For validation of the method at the 95 % confidence limit, the decision limits (CCα) determined from an extracted matrix curve were calculated to be 450, 36 and 24 μg/kg, and the detection capability (CCβ) as a screening method is ≤10 mg/kg, ≤160 μg/kg and ≤400 μg/kg for domoic acid, okadaic acid and saxitoxin, respectively.

A computational intelligent approach to multi-factor analysis of violent crime information systems

Various scientific studies have explored the causes of violent behaviour from different perspectives, with psychological tests, in particular, applied to the analysis of crime factors. The relationship between bi-factors has also been extensively studied including the link between age and crime. In reality, many factors interact to contribute to criminal behaviour and as such there is a need to have a greater level of insight into its complex nature. In this article we analyse violent crime information systems containing data on psychological, environmental and genetic factors. Our approach combines elements of rough set theory with fuzzy logic and particle swarm optimisation to yield an algorithm and methodology that can effectively extract multi-knowledge from information systems. The experimental results show that our approach outperforms alternative genetic algorithm and dynamic reduct-based techniques for reduct identification and has the added advantage of identifying multiple reducts and hence multi-knowledge (rules). Identified rules are consistent with classical statistical analysis of violent crime data and also reveal new insights into the interaction between several factors. As such, the results are helpful in improving our understanding of the factors contributing to violent crime and in highlighting the existence of hidden and intangible relationships between crime factors.

Correlation between ΔAbs, ΔRGB (red) and stearic acid destruction rates using commercial self-cleaning glass as the photocatalyst

A resazurin (Rz) based photocatalyst indicator ink is used to test the activity of a commercial self-cleaning glass, using UV–vis spectroscopy and digital photography to monitor the photocatalyst-driven change in colour of the ink. UV–vis spectroscopy allows the change in film absorbance, ΔAbs, to be monitored as a function of irradiation time, whereas digital photography is used to monitor the concomitant change in the red component of the RGB values, i.e. ΔRGB (red). Initial work reveals the variation in ΔAbst and ΔRGB (red)t as a function of irradiation time, t, are linearly correlated. The rates of change of these parameters are also linearly correlated to the rates of oxidative destruction of stearic acid on self-cleaning glass under different irradiances. This work demonstrates that a measure of photocatalyst activity of self-cleaning glass, i.e. the time taken to change the colour of an Rz photocatalyst indicator ink, can be obtained using inexpensive digital photography, as alternative to more expensive lab-based techniques, such as UV–vis spectrophotometry.

A bond model for DEM simulation of cementitious materials and deformable structures

There is an increasing use of the discrete element method (DEM) to study cemented (e.g. concrete and rocks) and sintered particulate materials. The chief advantage of the DEM over continuum based techniques is that it does not make assumptions about how cracking and fragmentation initiate and propagate, since the DEM system is naturally discontinuous. The ability for the DEM to produce a realistic representation of a cemented granular material depends largely on the implementation of an inter-particle bonded contact model. This paper presents a new bonded contact model based on the Timoshenko beam theory which considers axial, shear and bending behaviour of the bond. The bond model was first verified by simulating both the bending and dynamic response of a simply supported beam. The loading response of a concrete cylinder was then investigated and compared with the Eurocode equation prediction. The results show significant potential for the new model to produce satisfactory predictions for cementitious materials. A unique feature of this model is that it can also be used to accurately represent many deformable structures such as frames and shells, so that both particles and structures or deformable boundaries can be described in the same DEM framework. 

Improving E-Assessment Systems With A focus On Adaptive Feedback And Marking

Institutions involved in the provision of tertiary education across Europe are feeling the pinch. European universities, and other higher education (HE) institutions, must operate in a climate where the pressure of government spending cuts (Garben, 2012) is in stark juxtaposition to the EU’s strategy to drive forward and maintain a growth of student numbers in the sector (eurostat, 2015).

In order to remain competitive, universities and HE institutions are making ever-greater use of electronic assessment (E-Assessment) systems (Chatzigavriil et all, 2015; Ferrell, 2012). These systems are attractive primarily because they offer a cost-effect and scalable approach for assessment. In addition to scalability, they also offer reliability, consistency and impartiality; furthermore, from the perspective of a student they are most popular because they can offer instant feedback (Walet, 2012).

There are disadvantages, though.

First, feedback is often returned to a student immediately on competition of their assessment. While it is possible to disable the instant feedback option (this is often the case during an end of semester exam period when assessment scores must be can be ratified before release), however, this option tends to be a global ‘all on’ or ‘all off’ configuration option which is controlled centrally rather than configurable on a per-assessment basis.

If a formative in-term assessment is to be taken by multiple groups of
students, each at different times, this restriction means that answers to each question will be disclosed to the first group of students undertaking the assessment. As soon as the answers are released “into the wild” the academic integrity of the assessment is lost for subsequent student groups.

Second, the style of feedback provided to a student for each question is often limited to a simple ‘correct’ or ‘incorrect’ indicator. While this type of feedback has its place, it often does not provide a student with enough insight to improve their understanding of a topic that they did not answer correctly.

Most E-Assessment systems boast a wide range of question types including Multiple Choice, Multiple Response, Free Text Entry/Text Matching and Numerical questions. The design of these types of questions is often quite restrictive and formulaic, which has a knock-on effect on the quality of feedback that can be provided in each case.

Multiple Choice Questions (MCQs) are most prevalent as they are the most prescriptive and therefore most the straightforward to mark consistently. They are also the most amenable question types, which allow easy provision of meaningful, relevant feedback to each possible outcome chosen.
Text matching questions tend to be more problematic due to their free text entry nature. Common misspellings or case-sensitivity errors can often be accounted for by the software but they are by no means fool proof, as it is very difficult to predict in advance the range of possible variations on an answer that would be considered worthy of marks by a manual marker of a paper based equivalent of the same question.

Numerical questions are similarly restricted. An answer can be checked for accuracy or whether it is within a certain range of the correct answer, but unless it is a special purpose-built mathematical E-Assessment system the system is unlikely to have computational capability and so cannot, for example, account for “method marks” which are commonly awarded in paper-based marking.

From a pedagogical perspective, the importance of providing useful formative feedback to students at a point in their learning when they can benefit from the feedback and put it to use must not be understated (Grieve et all, 2015; Ferrell, 2012).

In this work, we propose a number of software-based solutions, which will overcome the limitations and inflexibilities of existing E-Assessment systems.

A Spiking Neural Network Model of the Medial Superior Olive using Spike Timing Dependent Plasticity for Sound Localisation

Sound localization can be defined as the ability to identify the position of an input sound source and is considered a powerful aspect of mammalian perception. For low frequency sounds, i.e., in the range 270 Hz-1.5 KHz, the mammalian auditory pathway achieves this by extracting the Interaural Time Difference between sound signals being received by the left and right ear. This processing is performed in a region of the brain known as the Medial Superior Olive (MSO). This paper presents a Spiking Neural Network (SNN) based model of the MSO. The network model is trained using the Spike Timing Dependent Plasticity learning rule using experimentally observed Head Related Transfer Function data in an adult domestic cat. The results presented demonstrate how the proposed SNN model is able to perform sound localization with an accuracy of 91.82% when an error tolerance of +/-10 degrees is used. For angular resolutions down to 2.5 degrees , it will be demonstrated how software based simulations of the model incur significant computation times. The paper thus also addresses preliminary implementation on a Field Programmable Gate Array based hardware platform to accelerate system performance.

Electrical characterization of metal-oxide-polymer devices for non-volatile memory applications

The objective of this thesis is to study the properties of resistive switching effect based on bistable resistive memory which is fabricated in the form of Al2O3/polymer diodes and to contribute to the elucidation of resistive switching mechanisms. Resistive memories were characterized using a variety of electrical techniques, including current-voltage measurements, small-signal impedance, and electrical noise based techniques. All the measurements were carried out over a large temperature range. Fast voltage ramps were used to elucidate the dynamic response of the memory to rapid varying electric fields. The temperature dependence of the current provided insight into the role of trapped charges in resistive switching. The analysis of fast current fluctuations using electric noise techniques contributed to the elucidation of the kinetics involved in filament formation/rupture, the filament size and correspondent current capabilities. The results reported in this thesis provide insight into a number of issues namely: (i) The fundamental limitations on the speed of operation of a bi-layer resistive memory are the time and voltage dependences of the switch-on mechanism. (ii) The results explain the wide spread in switching times reported in the literature and the apparently anomalous behaviour of the high conductance state namely the disappearance of the negative differential resistance region at high voltage scan rates which is commonly attributed to a “dead time” phenomenon which had remained elusive since it was first reported in the ‘60s. (iii) Assuming that the current is filamentary, Comsol simulations were performed and used to explain the observed dynamic properties of the current-voltage characteristics. Furthermore, the simulations suggest that filaments can interact with each other. (iv) The current-voltage characteristics have been studied as a function of temperature. The findings indicate that creation and annihilation of filaments is controlled by filling and neutralizing traps localized at the oxide/polymer interface. (v) Resistive switching was also studied in small-molecule OLEDs. It was shown that the degradation that leads to a loss of light output during operation is caused by the presence of a resistive switching layer. A diagnostic tool that predicts premature failure of OLEDs was devised and proposed. Resistive switching is a property of oxides. These layers can grow in a number of devices including, organic light emitting diodes (OLEDs), spin-valve transistors and photovoltaic devices fabricated in different types of material. Under strong electric fields the oxides can undergo dielectric breakdown and become resistive switching layers. Resistive switching strongly modifies the charge injection causing a number of deleterious effects and eventually device failure. In this respect the findings in this thesis are relevant to understand reliability issues in devices across a very broad field.