Evaluation of texture and geometry for dimensional facial expression recognition

Facial expression recognition (FER) algorithms mainly focus on classification into a small discrete set of emotions or representation of emotions using facial action units (AUs). Dimensional representation of emotions as continuous values in an arousal-valence space is relatively less investigated. It is not fully known whether fusion of geometric and texture features will result in better dimensional representation of spontaneous emotions. Moreover, the performance of many previously proposed approaches to dimensional representation has not been evaluated thoroughly on publicly available databases. To address these limitations, this paper presents an evaluation framework for dimensional representation of spontaneous facial expressions using texture and geometric features. SIFT, Gabor and LBP features are extracted around facial fiducial points and fused with FAP distance features. The CFS algorithm is adopted for discriminative texture feature selection. Experimental results evaluated on the publicly accessible NVIE database demonstrate that fusion of texture and geometry does not lead to a much better performance than using texture alone, but does result in a significant performance improvement over geometry alone. LBP features perform the best when fused with geometric features. Distributions of arousal and valence for different emotions obtained via the feature extraction process are compared with those obtained from subjective ground truth values assigned by viewers. Predicted valence is found to have a more similar distribution to ground truth than arousal in terms of covariance or Bhattacharya distance, but it shows a greater distance between the means.

Inexact uniformization method for computing transient distributions of Markov chains

The uniformization method (also known as randomization) is a numerically stable algorithm for computing transient distributions of a continuous time Markov chain. When the solution is needed after a long run or when the convergence is slow, the uniformization method involves a large number of matrix-vector products. Despite this, the method remains very popular due to its ease of implementation and its reliability in many practical circumstances. Because calculating the matrix-vector product is the most time-consuming part of the method, overall efficiency in solving large-scale problems can be significantly enhanced if the matrix-vector product is made more economical. In this paper, we incorporate a new relaxation strategy into the uniformization method to compute the matrix-vector products only approximately. We analyze the error introduced by these inexact matrix-vector products and discuss strategies for refining the accuracy of the relaxation while reducing the execution cost. Numerical experiments drawn from computer systems and biological systems are given to show that significant computational savings are achieved in practical applications.

The geometric modelling of social frames and contexts

How do humans respond to their social context? This question is becoming increasingly urgent in a society where democracy requires that the citizens of a country help to decide upon its policy directions, and yet those citizens frequently have very little knowledge of the complex issues that these policies seek to address. Frequently, we find that humans make their decisions more with reference to their social setting, than to the arguments of scientists, academics, and policy makers. It is broadly anticipated that the agent based modelling (ABM) of human behaviour will make it possible to treat such social effects, but we take the position here that a more sophisticated treatment of context will be required in many such models. While notions such as historical context (where the past history of an agent might affect its later actions) and situational context (where the agent will choose a different action in a different situation) abound in ABM scenarios, we will discuss a case of a potentially changing context, where social effects can have a strong influence upon the perceptions of a group of subjects. In particular, we shall discuss a recently reported case where a biased worm in an election debate led to significant distortions in the reports given by participants as to who won the debate (Davis et al 2011). Thus, participants in a different social context drew different conclusions about the perceived winner of the same debate, with associated significant differences among the two groups as to who they would vote for in the coming election. We extend this example to the problem of modelling the likely electoral responses of agents in the context of the climate change debate, and discuss the notion of interference between related questions that might be asked of an agent in a social simulation that was intended to simulate their likely responses. A modelling technology which could account for such strong social contextual effects would benefit regulatory bodies which need to navigate between multiple interests and concerns, and we shall present one viable avenue for constructing such a technology. A geometric approach will be presented, where the internal state of an agent is represented in a vector space, and their social context is naturally modelled as a set of basis states that are chosen with reference to the problem space.

Insect distributions and sampling protocols for stored commodities

Purpose of review: This review provides an overview on the importance of characterising and considering insect distribution infor- mation for designing stored commodity sampling protocols. Findings: Sampling protocols are influenced by a number of factors including government regulations, management practices, new technology and current perceptions of the status of insect pest damage. The spatial distribution of insects in stored commodities influ- ences the efficiency of sampling protocols; these can vary in response to season, treatment and other factors. It is important to use sam- pling designs based on robust statistics suitable for the purpose. Future research: The development of sampling protocols based on flexible, robust statistics allows for accuracy across a range of spatial distributions. Additionally, power can be added to sampling protocols through the integration of external information such as treatment history and climate. Bayesian analysis provides a coherent and well understood means to achieve this.

A mask-based approach for the geometric calibration of thermal-infrared cameras

Accurate and efficient thermal-infrared (IR) camera calibration is important for advancing computer vision research within the thermal modality. This paper presents an approach for geometrically calibrating individual and multiple cameras in both the thermal and visible modalities. The proposed technique can be used to correct for lens distortion and to simultaneously reference both visible and thermal-IR cameras to a single coordinate frame. The most popular existing approach for the geometric calibration of thermal cameras uses a printed chessboard heated by a flood lamp and is comparatively inaccurate and difficult to execute. Additionally, software toolkits provided for calibration either are unsuitable for this task or require substantial manual intervention. A new geometric mask with high thermal contrast and not requiring a flood lamp is presented as an alternative calibration pattern. Calibration points on the pattern are then accurately located using a clustering-based algorithm which utilizes the maximally stable extremal region detector. This algorithm is integrated into an automatic end-to-end system for calibrating single or multiple cameras. The evaluation shows that using the proposed mask achieves a mean reprojection error up to 78% lower than that using a heated chessboard. The effectiveness of the approach is further demonstrated by using it to calibrate two multiple-camera multiple-modality setups. Source code and binaries for the developed software are provided on the project Web site.

The development of ground impact models for the analysis of the risks associated with Unmanned Aircraft Operations over inhabited areas

With the emergence of Unmanned Aircraft Systems (UAS) there is a growing need for safety standards and regulatory frameworks to manage the risks associated with their operations. The primary driver for airworthiness regulations (i.e., those governing the design, manufacture, maintenance and operation of UAS) are the risks presented to people in the regions overflown by the aircraft. Models characterising the nature of these risks are needed to inform the development of airworthiness regulations. The output from these models should include measures of the collective, individual and societal risk. A brief review of these measures is provided. Based on the review, it was determined that the model of the operation of an UAS over inhabited areas must be capable of describing the distribution of possible impact locations, given a failure at a particular point in the flight plan. Existing models either do not take the impact distribution into consideration, or propose complex and computationally expensive methods for its calculation. A computationally efficient approach for estimating the boundary (and in turn area) of the impact distribution for fixed wing unmanned aircraft is proposed. A series of geometric templates that approximate the impact distributions are derived using an empirical analysis of the results obtained from a 6-Degree of Freedom (6DoF) simulation. The impact distributions can be aggregated to provide impact footprint distributions for a range of generic phases of flight and missions. The maximum impact footprint areas obtained from the geometric template are shown to have a relative error of typically less than 1% compared to the areas calculated using the computationally more expensive 6DoF simulation. Computation times for the geometric models are on the order of one second or less, using a standard desktop computer. Future work includes characterising the distribution of impact locations within the footprint boundaries.

Restructuring of carbonaceous particles upon exposure to organic vapours

Recent research has described the restructuring of particles upon exposure to organic vapours; however, as yet hypotheses able to explain this phenomenon are limited. In this study, a range of experiments were performed to explore different hypotheses related to carbonaceous particle restructuring upon exposure to organic and water vapours, such as: the effect of surface tension, the role of organics in flocculating primary particles, as well as the ability of vapours to “wet” the particle surface. The change in mobility diameter (dm) was investigated for a range carbonaceous particle types (diesel exhaust, petrol exhaust, cigarette smoke, candle smoke, particles generated in a heptane/toluene flame, and wood smoke particles) exposed to different organic (heptane, ethanol, and dimethyl sulfoxide/water (1:1 vol%) mixture) and water vapours. Particles were first size-selected and then bubbled through an impinger (bubbler) containing either an organic solvent or water, where particles trapped inside rising bubbles were exposed to saturated vapours of the solvent in the impinger. The size distribution of particles was simultaneously measured upstream and downstream from the impinger. A size-dependent reduction in dm was observed when bubbling diesel exhaust, particles generated in a heptane/toluene flame, and candle smoke particles through heptane, ethanol and a dimethyl sulfoxide/water (1:1 vol %) mixture. In addition, the size distributions of particles bubbled through an impinger were broader. Moreover, an increase of the geometric standard deviation (σ) of the size distributions of particles bubbled through an impinger was also found to be size-dependent. Size-dependent reduction in dm and an increase of σ indicate that particles undergo restructuring to a more compact form, which was confirmed by TEM analysis. However, bubbling of these particles through water did not result in a size-dependent reduction in dm, nor in an increase of σ. Cigarette smoke, petrol exhaust, and wood smoke particles did not result in any substantial change in dm, or σ, when bubbled through organic solvents or water. Therefore, size-dependent reduction in the dm upon bubbling through organic solvents was observed only for particles that had a fractal-like structure, whilst particles that were liquid or were assumed to be spherical did not exhibit any reduction in dm. Compaction of fractal-like particles was attributed to the ability of condensing vapours to efficiently wet the particles. Our results also show that the presence of an organic layer on the surface of fractal-like particles, or the surface tension of the condensed liquid do not influence the extent of compaction.

Accurate stratospheric particle size distributions from two separate flat-plate collection surfaces

Over the past two decades, flat-plate particle collections have revealed the presence of a remarkable variety of both terrestrial and extraterrestrial material in the stratosphere [1-6]. The ratio of terrestrial to extraterrestrial material and the nature of material collected may vary over observable time scales. Variations in particle number density can be important since the earth’s atmospheric radiation balance, and therefore the earth’s climate, can be influenced by articulate absorption and scattering of radiation from the sun and earth [7-9]. In order to assess the number density of solid particles in the stratosphere, we have examined a representative fraction of the so1id particles from two flat-plate collection surfaces, whose collection dates are separated in time by 5 years.

Grain size distributions of Magneli phases and metallic titanium in chondritic porous interplanetary dust particles

A mineralogical survey of chondritic interplanetary dust particles (IDPs)showed that these micrometeorites differ significantly in form and texture from components of carbonaceous chondrites and contain some mineral assemblages which do not occur in any meteorite class1. Models of chondritic IDP mineral evolution generally ignore the typical (ultra-) fine grain size of consituent minerals which range between 0.002-0.1µm in size2. The chondritic porous (CP) subset of chondritic IDPs is probably debris from short period comets although evidence for a cometary origin is still circumstantial3. If CP IDPs represent dust from regions of the Solar System in which comet accretion occurred, it can be argued that pervasive mineralogical evolution of IDP dust has been arrested due to cryogenic storage in comet nuclei. Thus, preservation in CP IDPs of "unusual meteorite minerals", such as oxides of tin, bismuth and titanium4, should not be dismissed casually. These minerals may contain specific information about processes that occurred in regions of the solar nebula, and early Solar System, which spawned the IDP parent bodies such as comets and C, P and D asteroids6. It is not fully appreciated that the apparent disparity between the mineralogy of CP IDPs and carbonaceous chondrite matrix may also be caused by the choice of electron-beam techniques with different analytical resolution. For example, Mg-Si-Fe distributions of Cl matrix obtained by "defocussed beam" microprobe analyses are displaced towards lower Fe-values when using analytical electron microscope (AEM)data which resolve individual mineral grains of various layer silicates and magnetite in the same matrix6,7. In general, "unusual meteorite minerals" in chondritic IDPs, such as metallic titanium, Tin01-n(Magneli phases) and anatase8 add to the mineral data base of fine-grained Solar System materials and provide constraints on processes that occurred in the early Solar System.

Accurate stratospheric particle-size distributions from a flat-plate collection surface

The first representative chemical, structural, and morphological analysis of the solid particles from a single collection surface has been performed. This collection surface sampled the stratosphere between 17 and 19km in altitude in the summer of 1981, and therefore before the 1982 eruptions of El Chichón. A particle collection surface was washed free of all particles with rinses of Freon and hexane, and the resulting wash was directed through a series of vertically stacked Nucleopore filters. The size cutoff for the solid particle collection process in the stratosphere is found to be considerably less than 1 μm. The total stratospheric number density of solid particles larger than 1μm in diameter at the collection time is calculated to be about 2.7×10−1 particles per cubic meter, of which approximately 95% are smaller than 5μm in diameter. Previous classification schemes are expanded to explicitly recognize low atomic number material. With the single exception of the calcium-aluminum-silicate (CAS) spheres all solid particle types show a logarithmic increase in number concentration with decreasing diameter. The aluminum-rich particles are unique in showing bimodal size distributions. In addition, spheres constitute only a minor fraction of the aluminum-rich material. About 2/3 of the particles examined were found to be shards of rhyolitic glass. This abundant volcanic material could not be correlated with any eruption plume known to have vented directly to the stratosphere. The micrometeorite number density calculated from this data set is 5×10−2 micrometeorites per cubic meter of air, an order of magnitude greater than the best previous estimate. At the collection altitude, the maximum collision frequency of solid particles >5μm in average diameter is calculated to be 6.91×10−16 collisions per second, which indicates negligible contamination of extraterrestrial particles in the stratosphere by solid anthropogenic particles.

Investigation of the airborne submicrometer particles emitted by dredging vessels using a plume capture method

This paper presents a method for investigating ship emissions, the plume capture and analysis system (PCAS), and its application in measuring airborne pollutant emission factors (EFs) and particle size distributions. The current investigation was conducted in situ, aboard two dredgers (Amity: a cutter suction dredger and Brisbane: a hopper suction dredger) but the PCAS is also capable of performing such measurements remotely at a distant point within the plume. EFs were measured relative to the fuel consumption using the fuel combustion derived plume CO2. All plume measurements were corrected by subtracting background concentrations sampled regularly from upwind of the stacks. Each measurement typically took 6 minutes to complete and during one day, 40 to 50 measurements were possible. The relationship between the EFs and plume sample dilution was examined to determine the plume dilution range over which the technique could deliver consistent results when measuring EFs for particle number (PN), NOx, SO2, and PM2.5 within a targeted dilution factor range of 50-1000 suitable for remote sampling. The EFs for NOx, SO2, and PM2.5 were found to be independent of dilution, for dilution factors within that range. The EF measurement for PN was corrected for coagulation losses by applying a time dependant particle loss correction to the particle number concentration data. For the Amity, the EF ranges were PN: 2.2 - 9.6 × 1015 (kg-fuel)-1; NOx: 35-72 g(NO2).(kg-fuel)-1, SO2 0.6 - 1.1 g(SO2).(kg-fuel)-1and PM2.5: 0.7 – 6.1 g(PM2.5).(kg-fuel)-1. For the Brisbane they were PN: 1.0 – 1.5 x 1016 (kg-fuel)-1, NOx: 3.4 – 8.0 g(NO2).(kg-fuel)-1, SO2: 1.3 – 1.7 g(SO2).(kg-fuel)-1 and PM2.5: 1.2 – 5.6 g(PM2.5).(kg-fuel)-1. The results are discussed in terms of the operating conditions of the vessels’ engines. Particle number emission factors as a function of size as well as the count median diameter (CMD), and geometric standard deviation of the size distributions are provided. The size distributions were found to be consistently uni-modal in the range below 500 nm, and this mode was within the accumulation mode range for both vessels. The representative CMDs for the various activities performed by the dredgers ranged from 94-131 nm in the case of the Amity, and 58-80 nm for the Brisbane. A strong inverse relationship between CMD and EF(PN) was observed.

A finite element model of seat cushion indentation with a soft tissue human occupant model

Effective digital human model (DHM) simulation of automotive driver packaging ergonomics, safety and comfort depends on accurate modelling of occupant posture, which is strongly related to the mechanical interaction between human body soft tissue and flexible seat components. This paper presents a finite-element study simulating the deflection of seat cushion foam and supportive seat structures, as well as human buttock and thigh soft tissue when seated. The three-dimensional data used for modelling thigh and buttock geometry were taken on one 95th percentile male subject, representing the bivariate percentiles of the combined hip breadth (seated) and buttock-to-knee length distributions of a selected Australian and US population. A thigh-buttock surface shell based on this data was generated for the analytic model. A 6mm neoprene layer was offset from the shell to account for the compression of body tissue expected through sitting in a seat. The thigh-buttock model is therefore made of two layers, covering thin to moderate thigh and buttock proportions, but not more fleshy sizes. To replicate the effects of skin and fat, the neoprene rubber layer was modelled as a hyperelastic material with viscoelastic behaviour in a Neo-Hookean material model. Finite element (FE) analysis was performed in ANSYS V13 WB (Canonsburg, USA). It is hypothesized that the presented FE simulation delivers a valid result, compared to a standard SAE physical test and the real phenomenon of human-seat indentation. The analytical model is based on the CAD assembly of a Ford Territory seat. The optimized seat frame, suspension and foam pad CAD data were transformed and meshed into FE models and indented by the two layer, soft surface human FE model. Converging results with the least computational effort were achieved for a bonded connection between cushion and seat base as well as cushion and suspension, no separation between neoprene and indenter shell and a frictional connection between cushion pad and neoprene. The result is compared to a previous simulation of an indentation with a hard shell human finite-element model of equal geometry, and to the physical indentation result, which is approached with very high fidelity. We conclude that (a) SAE composite buttock form indentation of a suspended seat cushion can be validly simulated in a FE model of merely similar geometry, but using a two-layer hard/soft structure. (b) Human-seat indentation of a suspended seat cushion can be validly simulated with a simplified human buttock-thigh model for a selected anthropomorphism.