An evaluation of crowd counting methods, features and regression models

Existing crowd counting algorithms rely on holistic, local or histogram based features to capture crowd properties. Regression is then employed to estimate the crowd size. Insufficient testing across multiple datasets has made it difficult to compare and contrast different methodologies. This paper presents an evaluation across multiple datasets to compare holistic, local and histogram based methods, and to compare various image features and regression models. A K-fold cross validation protocol is followed to evaluate the performance across five public datasets: UCSD, PETS 2009, Fudan, Mall and Grand Central datasets. Image features are categorised into five types: size, shape, edges, keypoints and textures. The regression models evaluated are: Gaussian process regression (GPR), linear regression, K nearest neighbours (KNN) and neural networks (NN). The results demonstrate that local features outperform equivalent holistic and histogram based features; optimal performance is observed using all image features except for textures; and that GPR outperforms linear, KNN and NN regression

Scene signatures : localised and point-less features for localisation

This paper is about localising across extreme lighting and weather conditions. We depart from the traditional point-feature-based approach as matching under dramatic appearance changes is a brittle and hard thing. Point feature detectors are fixed and rigid procedures which pass over an image examining small, low-level structure such as corners or blobs. They apply the same criteria applied all images of all places. This paper takes a contrary view and asks what is possible if instead we learn a bespoke detector for every place. Our localisation task then turns into curating a large bank of spatially indexed detectors and we show that this yields vastly superior performance in terms of robustness in exchange for a reduced but tolerable metric precision. We present an unsupervised system that produces broad-region detectors for distinctive visual elements, called scene signatures, which can be associated across almost all appearance changes. We show, using 21km of data collected over a period of 3 months, that our system is capable of producing metric localisation estimates from night-to-day or summer-to-winter conditions.

Speaker verification using I-vector features

This PhD research has provided novel solutions to three major challenges which have prevented the wide spread deployment of speaker recognition technology: (1) combating enrolment/ verification mismatch, (2) reducing the large amount of development and training data that is required and (3) reducing the duration of speech required to verify a speaker. A range of applications of speaker recognition technology from forensics in criminal investigations to secure access in banking will benefit from the research outcomes.

Multiscale features and information extraction of online strain for long-span bridges

The strain data acquired from structural health monitoring (SHM) systems play an important role in the state monitoring and damage identification of bridges. Due to the environmental complexity of civil structures, a better understanding of the actual strain data will help filling the gap between theoretical/laboratorial results and practical application. In the study, the multi-scale features of strain response are first revealed after abundant investigations on the actual data from two typical long-span bridges. Results show that, strain types at the three typical temporal scales of 10^5, 10^2 and 10^0 sec are caused by temperature change, trains and heavy trucks, and have their respective cut-off frequency in the order of 10^-2, 10^-1 and 10^0 Hz. Multi-resolution analysis and wavelet shrinkage are applied for separating and extracting these strain types. During the above process, two methods for determining thresholds are introduced. The excellent ability of wavelet transform on simultaneously time-frequency analysis leads to an effective information extraction. After extraction, the strain data will be compressed at an attractive ratio. This research may contribute to a further understanding of actual strain data of long-span bridges; also, the proposed extracting methodology is applicable on actual SHM systems.

An MRF based abnormal event detection approach using motion and appearance features

Abnormal event detection has attracted a lot of attention in the computer vision research community during recent years due to the increased focus on automated surveillance systems to improve security in public places. Due to the scarcity of training data and the definition of an abnormality being dependent on context, abnormal event detection is generally formulated as a data-driven approach where activities are modeled in an unsupervised fashion during the training phase. In this work, we use a Gaussian mixture model (GMM) to cluster the activities during the training phase, and propose a Gaussian mixture model based Markov random field (GMM-MRF) to estimate the likelihood scores of new videos in the testing phase. Further-more, we propose two new features: optical acceleration, and the histogram of optical flow gradients; to detect the presence of any abnormal objects and speed violations in the scene. We show that our proposed method outperforms other state of the art abnormal event detection algorithms on publicly available UCSD dataset.

Evaluation of features for leaf classification in challenging conditions

Fine-grained leaf classification has concentrated on the use of traditional shape and statistical features to classify ideal images. In this paper we evaluate the effectiveness of traditional hand-crafted features and propose the use of deep convolutional neural network (ConvNet) features. We introduce a range of condition variations to explore the robustness of these features, including: translation, scaling, rotation, shading and occlusion. Evaluations on the Flavia dataset demonstrate that in ideal imaging conditions, combining traditional and ConvNet features yields state-of-theart performance with an average accuracy of 97:3%�0:6% compared to traditional features which obtain an average accuracy of 91:2%�1:6%. Further experiments show that this combined classification approach consistently outperforms the best set of traditional features by an average of 5:7% for all of the evaluated condition variations.

Prevalence of malnutrition, obesity and nutritional risk of Australian paediatric inpatients: A national one-day snapshot

Aim Low prevalence rates of malnutrition at 2.5% to 4% have previously been reported in two tertiary paediatric Australian hospitals. The current study is the first to measure the prevalence of malnutrition, obesity and nutritional risk of paediatric inpatients in multiple hospitals throughout Australia. Methods Malnutrition, obesity and nutritional risk prevalence were investigated in 832 and 570 paediatric inpatients, respectively, in eight tertiary paediatric hospitals and eight regional hospitals across Australia on a single day. Malnutrition and obesity prevalence was determined using z-scores and body mass index (BMI) percentiles. High nutritional risk was determined as a Paediatric Yorkhill Malnutrition Score of 2 or more. Results The prevalence rates of malnourished, wasted, stunted, overweight and obese paediatric patients were 15%, 13.8%, 11.9%, 8.8% and 9.9%, respectively. Patients who identified as Aboriginal and Torres Strait Islander were more likely to have lower height-for-age z-scores (P < 0.01); however, BMI and weight-for-age z-scores were not significantly different. Children who were younger, from regional hospitals or with a primary diagnosis of cardiac disease or cystic fibrosis had significantly lower anthropometric z-scores (P = 0.05). Forty-four per cent of patients were identified as at high nutritional risk and requiring further nutritional assessment. Conclusions The prevalence of malnutrition and nutritional risk of Australian paediatric inpatients on a given day was much higher when compared with the healthy population. In contrast, the proportion of overweight and obese patients was less.

Improved nutritional status is related to improved quality of life in Parkinson’s disease

Background: Quality of life is poorer in Parkinson’s disease than in other conditions and in the general population without Parkinson’s disease. Malnutrition also results in poorer quality of life. This study aimed at determining the relationship between quality of life and nutritional status. Methods: Community-dwelling people with Parkinson’s disease >18 years old were recruited. The Patient-Generated Subjective Global Assessment (PG-SGA) assessed nutritional status. The Parkinson’s Disease Questionnaire 39 (PDQ-39) measured quality of life. Phase I was cross-sectional. The malnourished in Phase I were eligible for a nutrition intervention phase, randomised into 2 groups: standard care (SC) with provision of nutrition education materials only and intervention (INT) with individualised dietetic advice and regular weekly follow-up. Data were collected at baseline, 6 weeks, and 12 weeks. Results: Phase I consisted of 120 people who completed the PDQ-39. Phase II consisted of 9 in the SC group and 10 in the INT group. In Phase I, quality of life was poorer in the malnourished, particularly for mobility and activities of daily living domains. There was a significant correlation between PG-SGA and PDQ-39 scores (Phase I, rs = 0.445, p = .000; Phase II, rs = .426, p = .002). In Phase II, no significant difference in the PDQ-39 total or sub-scores was observed between the INT and SC groups; however, there was significant improvement in the emotional well-being domain for the entire group, X2(2) = 8.84, p = .012. Conclusions: Malnourished people with Parkinson’s disease had poorer quality of life than the well-nourished, and improvements in nutritional status resulted in quality of life improvements. Attention to nutritional status is an important component of quality of life and therefore the total care of people with Parkinson’s disease.

Nutritional support in chronic obstructive Pulmonary Disease (COPD) a randomised trial

Rationale Nutritional support is effective in managing malnutrition in COPD (Collins et al., 2012) leading to functional improvements (Collins et al., 2013). However, comparative trials of first line interventions are lacking. This randomised trial compared the effectiveness of individualised dietary advice by a dietitian (DA) versus oral nutritional supplements (ONS). Methods A target sample of 200 stable COPD outpatients at risk of malnutrition (‘MUST’; medium + high risk) were randomised to either a 12-week intervention of ONS (ONS: ~400 kcal/d, ~40 g/d protein) or DA with supportive written advice. The primary outcome was quality of life (QoL) measured using St George’s Respiratory Questionnaire with secondary outcomes including handgrip strength, body weight and nutritional intake. Both the change from baseline and the differences between groups was analysed using SPSS version 20. Results 84 outpatients were recruited (ONS: 41 vs. DA: 43), 72 completed the intervention (ONS: 33 vs. DA: 39). Mean BMI was 18.2 SD 1.6 kg/m2, age 72.6 SD 10 years, FEV1% predicted 36 SD 15% (severe COPD). In comparison to the DA group, the ONS group experienced significantly greater improvements in protein intakes above baseline values at both week 6 (+21.0 SEM 4.3 g/d vs. +0.52 SEM 4.3 g/d; p < 0.001) and week 12 (+19.0 SEM 5.0 g/d vs. +1.0 SEM 3.6 g/d; p = 0.033;ANOVA). QoL and secondary outcomes remained stable at 12 weeks in both groups with slight improvements in the ONS group but no differences between groups. Conclusion In outpatients at risk of malnutrition with severe COPD, nutritional support involving either ONS or DA appears to maintain in tritional status, functional capacity and QoL. However, larger trials, and earlier, multi-modal nutritional interventions for an extended duration should be explored.

The effectiveness of dietary counselling versus oral nutrition supplements in improving nutritional intake in malnourished patients with Chronic Obstructive Pulmonary Disease (COPD)

The evidence for nutritional support in COPD is almost entirely based on oral nutritional supplements (ONS) yet despite this dietary counseling and food fortification (DA) are often used as the first line treatment for malnutrition. This study aimed to investigate the effectiveness of ONS vs. DA in improving nutritional intake in malnourished outpatients with COPD. 70 outpatients (BMI 18.4 SD 1.6 kg/m2, age 73 SD 9 years, severe COPD) were randomised to receive a 12-week intervention of either ONS or DA (n 33 ONS vs. n 37 DA). Paired t-test analysis revealed total energy intakes significantly increased with ONS at week 6 (+302 SD 537 kcal/d; p = 0.002), with a slight reduction at week 12 (+243 SD 718 kcal/d; p = 0.061) returning to baseline levels on stopping supplementation. DA resulted in small increases in energy that only reached significance 3 months post-intervention (week 6: +48 SD 623 kcal/d, p = 0.640; week 12: +157 SD 637 kcal/d, p = 0.139; week 26: +247 SD 592 kcal/d, p = 0.032). Protein intake was significantly higher in the ONS group at both week 6 and 12 (ONS: +19.0 SD 25.0 g/d vs. DA: +1.0 SD 13.0 g/d; p = 0.033 ANOVA) but no differences were found at week 26. Vitamin C, Iron and Zinc intakes significantly increased only in the ONS group. ONS significantly increased energy, protein and several micronutrient intakes in malnourished COPD patients but only during the period of supplementation. Trials investigating the effects of combined nutritional interventions are required.

Achieving competence: The interactional features of children's storytelling

Early years researchers interested in storytelling have largely focused on the development of children’s language and social skills within constructed story sessions. Less focus has been given to the interactional aspects of storytelling in children’s everyday conversation and how the members themselves, the storytellers and story recipients, manage storytelling. An interactional view, using ethnomethodological and conversation analytic approaches, offers the opportunity to study children’s narratives in terms of ‘members work’. Detailed examination of a video-recorded interaction among a group of children in a preparatory year playground shows how the children managed interactions within conversational storytelling. Analyses highlight the ways in which children worked at gaining a turn and made a story tellable within a round of second stories. Investigating children’s competence-in-action ‘from within’, the findings from this research show how children invoke and accomplish competence through their interactions.

Identifying weak linear features with the "coalescing shortest path image transform"

The detection of line-like features in images finds many applications in microanalysis. Actin fibers, microtubules, neurites, pilis, DNA, and other biological structures all come up as tenuous curved lines in microscopy images. A reliable tracing method that preserves the integrity and details of these structures is particularly important for quantitative analyses. We have developed a new image transform called the "Coalescing Shortest Path Image Transform" with very encouraging properties. Our scheme efficiently combines information from an extensive collection of shortest paths in the image to delineate even very weak linear features. © Copyright Microscopy Society of America 2011.

Using upper bounds on attainable discrimination to select discrete valued features

Selection of features that will permit accurate pattern classification is a difficult task. However, if a particular data set is represented by discrete valued features, it becomes possible to determine empirically the contribution that each feature makes to the discrimination between classes. This paper extends the discrimination bound method so that both the maximum and average discrimination expected on unseen test data can be estimated. These estimation techniques are the basis of a backwards elimination algorithm that can be use to rank features in order of their discriminative power. Two problems are used to demonstrate this feature selection process: classification of the Mushroom Database, and a real-world, pregnancy related medical risk prediction task - assessment of risk of perinatal death.

Nutritional challenges for the family caregiver and person with dementia dyad

ABSTRACT Background: The majority of people with dementia live at home until quite late in the disease trajectory, supported by family caregivers who typically take increasing responsibility for providing nutrition. Caregiving is highly stressful and thus both dyad partners are at risk of nutritional issues. Objective: This study evaluated the nutritional status of both dyad members and the associations between these. Design Descriptive, correlational Setting Community Participants 26 dyads of persons with dementia and caregivers Measurements: The nutritional status of each dyad member was evaluated at home using a comprehensive battery of measures including the Mini-Nutritional Assessment, Corrected Arm Muscle Area and a 3-day food diary. Stage of dementia and functional eating capacity was measured for the person with dementia. Caregivers completed a brief burden scale. Results: Of those with dementia (n = 26), a large proportion had nutritional issues (one was malnourished and another 16 were at risk). Six of the caregivers were at risk of malnutrition. In addition, fifteen of the people with dementia did not meet their recommended daily energy requirements. A moderate and significant positive correlation between functional eating skills and nutritional status (MNA score) among participants with dementia was found (r =.523, n = 26, p.006). Conclusion: These findings suggest that a dyadic perspective of nutritional status provides important insights into risk in this vulnerable group. Specifically, monitoring of the functional eating independence skills of the person with dementia is critical, along with assisting caregivers to be aware of their own eating patterns and intake.