Gradual training of cascaded shape regression for facial landmark localization and pose estimation

Facial landmarks play an important role in face recognition. They serve different steps of the recognition such as pose estimation, face alignment, and local feature extraction. Recently, cascaded shape regression has been proposed to accurately locate facial landmarks. A large number of weak regressors are cascaded in a sequence to fit face shapes to the correct landmark locations. In this paper, we propose to improve the method by applying gradual training. With this training, the regressors are not directly aimed to the true locations. The sequence instead is divided into successive parts each of which is aimed to intermediate targets between the initial and the true locations. We also investigate the incorporation of pose information in the cascaded model. The aim is to find out whether the model can be directly used to estimate head pose. Experiments on the Annotated Facial Landmarks in the Wild database have shown that the proposed method is able to improve the localization and give accurate estimates of pose.

How long does it take for aquifer recharge or aquifer discharge processes to reach steady state?

Groundwater flow models are usually characterized as being either transient flow models or steady state flow models. Given that steady state groundwater flow conditions arise as a long time asymptotic limit of a particular transient response, it is natural for us to seek a finite estimate of the amount of time required for a particular transient flow problem to effectively reach steady state. Here, we introduce the concept of mean action time (MAT) to address a fundamental question: How long does it take for a groundwater recharge process or discharge processes to effectively reach steady state? This concept relies on identifying a cumulative distribution function, $F(t;x)$, which varies from $F(0;x)=0$ to $F(t;x) \to \infty$ as $t\to \infty$, thereby providing us with a measurement of the progress of the system towards steady state. The MAT corresponds to the mean of the associated probability density function $f(t;x) = \dfrac{dF}{dt}$, and we demonstrate that this framework provides useful analytical insight by explicitly showing how the MAT depends on the parameters in the model and the geometry of the problem. Additional theoretical results relating to the variance of $f(t;x)$, known as the variance of action time (VAT), are also presented. To test our theoretical predictions we include measurements from a laboratory–scale experiment describing flow through a homogeneous porous medium. The laboratory data confirms that the theoretical MAT predictions are in good agreement with measurements from the physical model.

Quantitative genetic parameter estimates for body and carcass traits in a cultured stock of giant freshwater prawn (Macrobrachium rosenbergii) selected for harvest weight in Vietnam

We estimated the heritability and correlations between body and carcass weight traits in a cultured stock of giant freshwater prawn (GFP) (Macrobrachium rosenbergii) selected for harvest body weight in Vietnam. The data set consisted of 18,387 body and 1,730 carcass records, as well as full pedigree information collected over four generations. Variance and covariance components were estimated by restricted maximum likelihood fitting a multi-trait animal model. Across generations, estimates of heritability for body and carcass weight traits were moderate and ranged from 0.14 to 0.19 and 0.17 to 0.21, respectively. Body trait heritabilities estimated for females were significantly higher than for males whereas carcass weight trait heritabilities estimated for females and males were not significantly different (P>. 0.05). Maternal effects for body traits accounted for 4 to 5% of the total variance and were greater in females than in males. Genetic correlations among body traits were generally high in the mixed sexes. Genetic correlations between body and carcass weight traits were also high. Although some issues remain regarding the best statistical model to be fitted to GFP data, our results suggest that selection for high harvest body weight based on breeding values estimated by fitting an animal model to the data can significantly improve mean body and carcass weight in GFP.

Traffic density estimation of signalised arterials with stop line detector and probe data

This research aims to develop a reliable density estimation method for signalised arterials based on cumulative counts from upstream and downstream detectors. In order to overcome counting errors associated with urban arterials with mid-link sinks and sources, CUmulative plots and Probe Integration for Travel timE estimation (CUPRITE) is employed for density estimation. The method, by utilizing probe vehicles’ samples, reduces or cancels the counting inconsistencies when vehicles’ conservation is not satisfied within a section. The method is tested in a controlled environment, and the authors demonstrate the effectiveness of CUPRITE for density estimation in a signalised section, and discuss issues associated with the method.

Real time information provision benefit measured by macroscopic fundamental diagram

The existence of Macroscopic Fundamental Diagram (MFD), which relates space-mean density and flow, has been shown in urban networks under homogeneous traffic conditions. Since MFD represents the area-wide network traffic performances, studies on perimeter control strategies and an area traffic state estimation utilizing the MFD concept has been reported. One of the key requirements for well-defined MFD is the homogeneity of the area-wide traffic condition with links of similar properties, which is not universally expected in real world. For the practical application of the MFD concept, several researchers have identified the influencing factors for network homogeneity. However, they did not explicitly take the impact of drivers’ behaviour and information provision into account, which has a significant impact on simulation outputs. This research aims to demonstrate the effect of dynamic information provision on network performance by employing the MFD as a measurement. A microscopic simulation, AIMSUN, is chosen as an experiment platform. By changing the ratio of en-route informed drivers and pre-trip informed drivers different scenarios are simulated in order to investigate how drivers’ adaptation to the traffic congestion influences the network performance with respect to the MFD shape as well as other indicators, such as total travel time. This study confirmed the impact of information provision on the MFD shape, and addressed the usefulness of the MFD for measuring the dynamic information provision benefit.

Genetic response to combined family selection for improved mean harvest weight in giant freshwater prawn (Macrobrachium rosenbergii) in Vietnam

We estimated genetic changes in body and carcass weight traits in a giant freshwater prawn (GFP) (Macrobrachium rosenbergii) population selected for increased body weight at harvest in Vietnam. The data set consisted of 18,387 individual body and 1730 carcass weight records, as well as full pedigree information collected over four generations. Average selection response (per generation) in body weight at harvest (transformed to square root) estimated as the difference between the Selection line and the Control group was 7.4% calculated from least squares mean (LSMs), 7.0% from estimated breeding values (EBVs) and 4.4% calculated from EBVs between two consecutive generations. Favorable correlated selection responses (estimated from LSMs) were found for other body traits including: total length, cephalothorax length, abdominal length, cephalothorax width, and abdominal width (12.1%, 14.5%, 10.4%, 15.5% and 13.3% over three selection generations, respectively). Data in the second generation of selection showed positive correlated responses for carcass weight traits including: abdominal weight, exoskeleton-off weight, and telson-off weight of 8.8%, 8.6% and 8.8%, respectively. We conclude that body weight at harvest responded well to the application of combined (between and within) family selection and correlated responses in carcass weight traits were favorable.

Econometric Modelling with Time Series : Specification, Estimation and Testing

This book provides a general framework for specifying, estimating, and testing time series econometric models. Special emphasis is given to estimation by maximum likelihood, but other methods are also discussed, including quasi-maximum likelihood estimation, generalized method of moments estimation, nonparametric estimation, and estimation by simulation. An important advantage of adopting the principle of maximum likelihood as the unifying framework for the book is that many of the estimators and test statistics proposed in econometrics can be derived within a likelihood framework, thereby providing a coherent vehicle for understanding their properties and interrelationships. In contrast to many existing econometric textbooks, which deal mainly with the theoretical properties of estimators and test statistics through a theorem-proof presentation, this book squarely addresses implementation to provide direct conduits between the theory and applied work.

Empirical evaluation of brisbane macroscopic fundamental diagram

The Macroscopic Fundamental Diagram (MFD) relates space-mean density and flow, and the existence with dynamic features was confirmed in congested urban network with real data set from loop detectors and taxi probes. Since the MFD represents the area-wide network traffic performances, it gives foundations for perimeter control strategies and an area traffic state estimation enabling area-based network control. However, limited works have been reported on real world example from signalised arterial network. This paper fuses data from multiple sources (Bluetooth, Loops and Signals) and develops a framework for the development of the MFD for Brisbane. Existence of the MFD in Brisbane network is confirmed. Different MFDs (from whole network and several sub regions) are evaluated to discover the spatial partitioning in network performance representation.

Sequential decision fusion of multibiometrics applied to text-dependent speaker verification for controlled errors

Reliability of the performance of biometric identity verification systems remains a significant challenge. Individual biometric samples of the same person (identity class) are not identical at each presentation and performance degradation arises from intra-class variability and inter-class similarity. These limitations lead to false accepts and false rejects that are dependent. It is therefore difficult to reduce the rate of one type of error without increasing the other. The focus of this dissertation is to investigate a method based on classifier fusion techniques to better control the trade-off between the verification errors using text-dependent speaker verification as the test platform. A sequential classifier fusion architecture that integrates multi-instance and multisample fusion schemes is proposed. This fusion method enables a controlled trade-off between false alarms and false rejects. For statistically independent classifier decisions, analytical expressions for each type of verification error are derived using base classifier performances. As this assumption may not be always valid, these expressions are modified to incorporate the correlation between statistically dependent decisions from clients and impostors. The architecture is empirically evaluated by applying the proposed architecture for text dependent speaker verification using the Hidden Markov Model based digit dependent speaker models in each stage with multiple attempts for each digit utterance. The trade-off between the verification errors is controlled using the parameters, number of decision stages (instances) and the number of attempts at each decision stage (samples), fine-tuned on evaluation/tune set. The statistical validation of the derived expressions for error estimates is evaluated on test data. The performance of the sequential method is further demonstrated to depend on the order of the combination of digits (instances) and the nature of repetitive attempts (samples). The false rejection and false acceptance rates for proposed fusion are estimated using the base classifier performances, the variance in correlation between classifier decisions and the sequence of classifiers with favourable dependence selected using the 'Sequential Error Ratio' criteria. The error rates are better estimated by incorporating user-dependent (such as speaker-dependent thresholds and speaker-specific digit combinations) and class-dependent (such as clientimpostor dependent favourable combinations and class-error based threshold estimation) information. The proposed architecture is desirable in most of the speaker verification applications such as remote authentication, telephone and internet shopping applications. The tuning of parameters - the number of instances and samples - serve both the security and user convenience requirements of speaker-specific verification. The architecture investigated here is applicable to verification using other biometric modalities such as handwriting, fingerprints and key strokes.

Vision-based estimation of airborne target pseudobearing rate using hidden Markov model filters

The problem of estimating pseudobearing rate information of an airborne target based on measurements from a vision sensor is considered. Novel image speed and heading angle estimators are presented that exploit image morphology, hidden Markov model (HMM) filtering, and relative entropy rate (RER) concepts to allow pseudobearing rate information to be determined before (or whilst) the target track is being estimated from vision information.

Incorporating spatial correlations into multispecies mean-field models

In biology, we frequently observe different species existing within the same environment. For example, there are many cell types in a tumour, or different animal species may occupy a given habitat. In modelling interactions between such species, we often make use of the mean field approximation, whereby spatial correlations between the locations of individuals are neglected. Whilst this approximation holds in certain situations, this is not always the case, and care must be taken to ensure the mean field approximation is only used in appropriate settings. In circumstances where the mean field approximation is unsuitable we need to include information on the spatial distributions of individuals, which is not a simple task. In this paper we provide a method that overcomes many of the failures of the mean field approximation for an on-lattice volume-excluding birth-death-movement process with multiple species. We explicitly take into account spatial information on the distribution of individuals by including partial differential equation descriptions of lattice site occupancy correlations. We demonstrate how to derive these equations for the multi-species case, and show results specific to a two-species problem. We compare averaged discrete results to both the mean field approximation and our improved method which incorporates spatial correlations. We note that the mean field approximation fails dramatically in some cases, predicting very different behaviour from that seen upon averaging multiple realisations of the discrete system. In contrast, our improved method provides excellent agreement with the averaged discrete behaviour in all cases, thus providing a more reliable modelling framework. Furthermore, our method is tractable as the resulting partial differential equations can be solved efficiently using standard numerical techniques.

A simplified approach for calculating the moments of action for linear reaction-diffusion equations

The mean action time is the mean of a probability density function that can be interpreted as a critical time, which is a finite estimate of the time taken for the transient solution of a reaction-diffusion equation to effectively reach steady state. For high-variance distributions, the mean action time under-approximates the critical time since it neglects to account for the spread about the mean. We can improve our estimate of the critical time by calculating the higher moments of the probability density function, called the moments of action, which provide additional information regarding the spread about the mean. Existing methods for calculating the nth moment of action require the solution of n nonhomogeneous boundary value problems which can be difficult and tedious to solve exactly. Here we present a simplified approach using Laplace transforms which allows us to calculate the nth moment of action without solving this family of boundary value problems and also without solving for the transient solution of the underlying reaction-diffusion problem. We demonstrate the generality of our method by calculating exact expressions for the moments of action for three problems from the biophysics literature. While the first problem we consider can be solved using existing methods, the second problem, which is readily solved using our approach, is intractable using previous techniques. The third problem illustrates how the Laplace transform approach can be used to study coupled linear reaction-diffusion equations.

Traffic density estimation of signalised arterials with stop line detector and probe data

This research aims to develop a reliable density estimation method for signalised arterials based on cumulative counts from upstream and downstream detectors. In order to overcome counting errors associated with urban arterials with mid-link sinks and sources, CUmulative plots and Probe Integration for Travel timE estimation (CUPRITE) is employed for density estimation. The method, by utilizing probe vehicles’ samples, reduces or cancels the counting inconsistencies when vehicles’ conservation is not satisfied within a section. The method is tested in a controlled environment, and the authors demonstrate the effectiveness of CUPRITE for density estimation in a signalised section, and discuss issues associated with the method.

Validity of dietary questionnaires in Sri Lankan adults and the association of dietary intake with obesity

Traditionally, infectious diseases and under-nutrition have been considered major health problems in Sri Lanka with little attention paid to obesity and associated non-communicable diseases (NCDs). However, the recent Sri Lanka Diabetes and Cardiovascular Study (SLDCS) reported the epidemic level of obesity, diabetes and metabolic syndrome. Moreover, obesity-associated NCDs is the leading cause of death in Sri Lanka and there is an exponential increase in hospitalization due to NCDs adversely affecting the development of the country. Despite Sri Lanka having a very high prevalence of NCDs and associated mortality, little is known about the causative factors for this burden. It is widely believed that the global NCD epidemic is associated with recent lifestyle changes, especially dietary factors. In the absence of sufficient data on dietary habits in Sri Lanka, successful interventions to manage these serious health issues would not be possible. In view of the current situation the dietary survey was undertaken to assess the intakes of energy, macro-nutrients and selected other nutrients with respect to socio demographic characteristics and the nutritional status of Sri Lankan adults especially focusing on obesity. Another aim of this study was to develop and validate a culturally specific food frequency questionnaire (FFQ) to assess dietary risk factors of NCDs in Sri Lankan adults. Data were collected from a subset of the national SLDCS using a multi-stage, stratified, random sampling procedure (n=500). However, data collection in the SLDCS was affected by the prevailing civil war which resulted in no data being collected from Northern and Eastern provinces. To obtain a nationally representative sample, additional subjects (n=100) were later recruited from the two provinces using similar selection criteria. Ethical Approval for this study was obtained from the Ethical Review Committee, Faculty of Medicine, University of Colombo, Sri Lanka and informed consent was obtained from the subjects before data were collected. Dietary data were obtained using the 24-h Dietary Recall (24HDR) method. Subjects were asked to recall all foods and beverages, consumed over the previous 24-hour period. Respondents were probed for the types of foods and food preparation methods. For the FFQ validation study, a 7-day weight diet record (7-d WDR) was used as the reference method. All foods recorded in the 24 HDR were converted into grams and then intake of energy and nutrients were analysed using NutriSurvey 2007 (EBISpro, Germany) which was modified for Sri Lankan food recipes. Socio-demographic details and body weight perception were collected from interviewer-administrated questionnaire. BMI was calculated and overweight (BMI ≥23 kg.m-2), obesity (BMI ≥25 kg.m-2) and abdominal obesity (Men: WC ≥ 90 cm; Women: WC ≥ 80 cm) were categorized according to Asia-pacific anthropometric cut-offs. The SPSS v. 16 for Windows and Minitab v10 were used for statistical analysis purposes. From a total of 600 eligible subjects, 491 (81.8%) participated of whom 34.5% (n=169) were males. Subjects were well distributed among different socio-economic parameters. A total of 312 different food items were recorded and nutritionists grouped similar food items which resulted in a total of 178 items. After performing step-wise multiple regression, 93 foods explained 90% of the variance for total energy intake, carbohydrates, protein, total fat and dietary fibre. Finally, 90 food items and 12 photographs were selected. Seventy-seven subjects completed (response rate = 65%) the FFQ and 7-day WDR. Estimated mean energy intake (SD) from FFQ (1794±398 kcal) and 7DWR (1698±333 kcal, P<0.001) was significantly different due to a significant overestimation of carbohydrate (~10 g/d, P<0.001) and to some extent fat (~5 g/d, NS). Significant positive correlations were found between the FFQ and 7DWR for energy (r = 0.39), carbohydrate (r = 0.47), protein (r = 0.26), fat (r =0.17) and dietary fiber (r = 0.32). Bland-Altman graphs indicated fairly good agreement between methods with no relationship between bias and average intake of each nutrient examined. The findings from the nutrition survey showed on average, Sri Lankan adults consumed over 14 portions of starch/d; moreover, males consumed 5 more portions of cereal than females. Sri Lankan adults consumed on average 3.56 portions of added sugars/d. Moreover, mean daily intake of fruit (0.43) and vegetable (1.73) portions was well below minimum dietary recommendations (fruits 2 portions/d; vegetables 3 portions/d). The total fruit and vegetable intake was 2.16 portions/d. Daily consumption of meat or alternatives was 1.75 portions and the sum of meat and pulses was 2.78 portions/d. Starchy foods were consumed by all participants and over 88% met the minimum daily recommendations. Importantly, nearly 70% of adults exceeded the maximum daily recommendation for starch (11portions/d) and a considerable proportion consumed larger numbers of starch servings daily, particularly men. More than 12% of men consumed over 25 starch servings/d. In contrast to their starch consumption, participants reported very low intakes of other food groups. Only 11.6%, 2.1% and 3.5% of adults consumed the minimum daily recommended servings of vegetables, fruits, and fruits and vegetables combined, respectively. Six out of ten adult Sri Lankans sampled did not consume any fruits. Milk and dairy consumption was extremely low; over a third of the population did not consume any dairy products and less than 1% of adults consumed 2 portions of dairy/d. A quarter of Sri Lankans did not report consumption of meat and pulses. Regarding protein consumption, 36.2% attained the minimum Sri Lankan recommendation for protein; and significantly more men than women achieved the recommendation of ≥3 servings of meat or alternatives daily (men 42.6%, women 32.8%; P<0.05). Over 70% of energy was derived from carbohydrates (Male:72.8±6.4%, Female:73.9±6.7%), followed by fat (Male:19.9±6.1%, Female:18.5±5.7%) and proteins (Male:10.6±2.1%, Female:10.9±5.6%). The average intake of dietary fiber was 21.3 g/day and 16.3 g/day for males and females, respectively. There was a significant difference in nutritional intake related to ethnicities, areas of residence, education levels and BMI categories. Similarly, dietary diversity was significantly associated with several socio-economic parameters among Sri Lankan adults. Adults with BMI ≥25 kg.m-2 and abdominally obese Sri Lankan adults had the highest diet diversity values. Age-adjusted prevalence (95% confidence interval) of overweight, obesity, and abdominal obesity among Sri Lankan adults were 17.1% (13.8-20.7), 28.8% (24.8-33.1), and 30.8% (26.8-35.2), respectively. Men, compared with women, were less overweight, 14.2% (9.4-20.5) versus 18.5% (14.4-23.3), P = 0.03, less obese, 21.0% (14.9-27.7) versus 32.7% (27.6-38.2), P < .05; and less abdominally obese, 11.9% (7.4-17.8) versus 40.6% (35.1-46.2), P < .05. Although, prevalence of obesity has reached to epidemic level body weight misperception was common among Sri Lankan adults. Two-thirds of overweight males and 44.7% of females considered themselves as in "about right weight". Over one third of both male and female obese subjects perceived themselves as "about right weight" or "underweight". Nearly 32% of centrally obese men and women perceived that their waist circumference is about right. People who perceived overweight or very overweight (n = 154) only 63.6% tried to lose their body weight (n = 98), and quarter of adults seek advices from professionals (n = 39). A number of important conclusions can be drawn from this research project. Firstly, the newly developed FFQ is an acceptable tool for assessing the nutrient intake of Sri Lankans and will assist proper categorization of individuals by dietary exposure. Secondly, a substantial proportion of the Sri Lankan population does not consume a varied and balanced diet, which is suggestive of a close association between the nutrition-related NCDs in the country and unhealthy eating habits. Moreover, dietary diversity is positively associated with several socio-demographic characteristics and obesity among Sri Lankan adults. Lastly, although obesity is a major health issue among Sri Lankan adults, body weight misperception was common among underweight, healthy weight, overweight, and obese adults in Sri Lanka. Over 2/3 of overweight and 1/3 of obese Sri Lankan adults believe that they are in "right weight" or "under-weight" categories.

Estimation of disability weights on malignant neoplasms in Shandong province [山东省现患恶性肿瘤残疾权重的测量]

Objective To determine stage-specific and average disability weights (DWs) of malignant neoplasm and provide support and evidence for study on burden of cancer and policy development in Shandong province. Methods Health status of each cancer patient identified during the cancer prevalence survey in Shandong, 2007 was investigated. In line with the GBD methodology in estimating DWs, the disability extent of every case was classified and evaluated according to the Six-class Disability Classification version and then the stage-specific weights and average DWs with their 95 % confidence intervals were calculated, using SAS software. Results A total of 11 757 cancer cases were investigated and evaluated. DWs of specific stage of therapy, remission, metastasis and terminal of all cancers were 0.310, 0.218, 0.450 and 0.653 respectively. The average DW of all cancers was 0.317(95 % CI:0.312-0.321). Weights of different stage and different cancer varied significantly, while no significant differences were found between males and females. DWs were found higher (>0.4) for liver cancer, bone cancer, lymphoma and pancreas cancer. Lower DWs (<0.3) were found for breast cancer, cervix uteri, corpus uteri, ovarian cancer, larynx cancer, mouth and oropharynx cancer. Conclusion Stage-specific and average DWs for various cancers were estimated based on a large sample size survey. The average DWs of 0.317 for all cancers indicated that 1/3 healthy year lost for each survived life year of them. The difference of DWs between different cancer and stage provide scientific evidence for cancer prevention strategy development. Abstract in Chinese 目的 测算各种恶性肿瘤的分病程残疾权重和平均残疾权重,为山东省恶性肿瘤疾病负担研究及肿瘤防治对策制定提供参考依据. 方法 在山东省2007年恶性肿瘤现患调查中对所有恶性肿瘤患者的健康状况进行调查,参考全球疾病负担研究的方法 ,利用六级社会功能分级标准对患者残疾状况进行分级和赋值,分别计算20种恶性肿瘤的分病程残疾权重和平均残疾权重及其95%CI. 结果 共调查恶性肿瘤患者11757例,所有恶性肿瘤治疗期、恢复期、转移期和晚期的残疾权重分别为0.310、0.218、0.450和0.653,平均残疾权重为0.317(95%CI:0.312～0.321).不同恶性肿瘤和不同病程阶段的残疾权重差别显著,性别间差异无统计学意义.肝癌、骨癌、淋巴瘤和胰腺癌平均残疾权重较高(>0.4),乳腺癌、子宫体癌、子宫颈癌、卵巢癌、喉癌和口咽部癌症相对较低(<0.3). 结论 山东省恶性肿瘤平均残疾权重为0.317,即恶性肿瘤患者每存活1年平均损失近1/3个健康生命年;不同恶性肿瘤和不同病程阶段的残疾权重差别为肿瘤防治对策的制定具有重要意义.