Depth sensor based 3D human motion estimation hardware such as Kinect has made interactive applications more popular recently. However, it is still challenging to accurately recognize postures from a ...single depth camera due to the inherently noisy data derived from depth images and self-occluding action performed by the user. In this paper, we propose a new real-time probabilistic framework to enhance the accuracy of live captured postures that belong to one of the action classes in the database. We adopt the Gaussian Process model as a prior to leverage the position data obtained from Kinect and marker-based motion capture system. We also incorporate a temporal consistency term into the optimization framework to constrain the velocity variations between successive frames. To ensure that the reconstructed posture resembles the accurate parts of the observed posture, we embed a set of joint reliability measurements into the optimization framework. A major drawback of Gaussian Process is its cubic learning complexity when dealing with a large database due to the inverse of a covariance matrix. To solve the problem, we propose a new method based on a local mixture of Gaussian Processes, in which Gaussian Processes are defined in local regions of the state space. Due to the significantly decreased sample size in each local Gaussian Process, the learning time is greatly reduced. At the same time, the prediction speed is enhanced as the weighted mean prediction for a given sample is determined by the nearby local models only. Our system also allows incrementally updating a specific local Gaussian Process in real time, which enhances the likelihood of adapting to run-time postures that are different from those in the database. Experimental results demonstrate that our system can generate high quality postures even under severe self-occlusion situations, which is beneficial for real-time applications such as motion-based gaming and sport training.
Makeup face verification in the wild is an important research problem for its popularization in real-world. However, little effort has been made to tackle it in computer vision. In this research, we ...first build a new database, i.e., Facial Beauty Database (FBD), which contains paired facial images of 8933 subjects without and with makeup in different real-world scenarios. To the best of our knowledge, FBD is the largest makeup face database to date compared with existing databases for facial makeup research. Moreover, we propose a new discriminative marginal metric learning (DMML) algorithm to deal with this problem in the wild. Inspired by the fact that interclass marginal faces are usually more discriminative than interclass nonmarginal faces in learning the discriminative metric space, we use the interclass marginal faces to depict the discriminative information. Simultaneously, we wish that those interclass marginal faces without makeup relations are separated from each other as far as possible, so that more discriminative information between facial images without and with makeup can be exploited for verification. Furthermore, since multiple features could provide comprehensive information in describing the facial representations from diverse points of view and extract more informative cues from facial images, we also introduce a multiview discriminative marginal metric learning (MDMML) algorithm by effectively learning a robust metric space such that multiple features from different points of view can be integrated to effectively enhance the performance of makeup face verification. Experimental results on two real-world makeup face databases are utilized to show the effectiveness of our method and the possibility of verifying the makeup relations from facial images in real-world.
The pursuit of early diagnosis of cerebral palsy has been an active research area with some very promising results using tools such as the General Movements Assessment (GMA). In our previous work, we ...explored the feasibility of extracting pose-based features from video sequences to automatically classify infant body movement into two categories, normal and abnormal. The classification was based upon the GMA, which was carried out on the video data by an independent expert reviewer. In this paper we extend our previous work by extracting the normalised pose-based feature sets, Histograms of Joint Orientation 2D (HOJO2D) and Histograms of Joint Displacement 2D (HOJD2D), for use in new deep learning architectures. We explore the viability of using these pose-based feature sets for automated classification within a deep learning framework by carrying out extensive experiments on five new deep learning architectures. Experimental results show that the proposed fully connected neural network FCNet performed robustly across different feature sets. Furthermore, the proposed convolutional neural network architectures demonstrated excellent performance in handling features in higher dimensionality. We make the code, extracted features and associated GMA labels publicly available.
For epidemiological and surveillance purposes, it is relevant to monitor the distribution and dynamics of Streptococcus pneumoniae serotypes. Conventional serotyping methods do not provide rapid or ...quantitative information on serotype loads. Quantitative serotyping may enable prediction of the invasiveness of a specific serotype compared to other serotypes carried. Here, we describe a novel, rapid multiplex real-time PCR assay for identification and quantification of the 40 most prevalent pneumococcal serotypes and the assay impacts in pneumonia specimens from emerging and developing countries. Eleven multiplex PCR to detect 40 serotypes or serogroups were optimized. Quantification was enabled by reference to standard dilutions of known bacterial load. Performance of the assay was evaluated to specifically type and quantify S. pneumoniae in nasopharyngeal and blood samples from adult and pediatric patients hospitalized with pneumonia (n = 664) from five different countries. Serogroup 6 was widely represented in nasopharyngeal specimens from all five cohorts. The most frequent serotypes in the French, South African, and Brazilian cohorts were 1 and 7A/F, 3 and 19F, and 14, respectively. When both samples were available, the serotype in blood was always present as carriage with other serotypes in the nasopharynx. Moreover, the ability of a serotype to invade the bloodstream may be linked to its nasopharyngeal load. The mean nasopharyngeal concentration of the serotypes that moved to the blood was 3 log-fold higher than the ones only found in the nasopharynx. This novel, rapid, quantitative assay may potentially predict some of the S. pneumoniae serotypes invasiveness and assessment of pneumococcal serotype distribution.
Real-Time Posture Reconstruction for Microsoft Kinect Shum, Hubert P. H.; Ho, Edmond S. L.; Jiang, Yang ...
IEEE transactions on cybernetics,
2013-Oct., 2013-Oct, 2013-10-00, 20131001, Letnik:
43, Številka:
5
Journal Article
Recenzirano
The recent advancement of motion recognition using Microsoft Kinect stimulates many new ideas in motion capture and virtual reality applications. Utilizing a pattern recognition algorithm, Kinect can ...determine the positions of different body parts from the user. However, due to the use of a single-depth camera, recognition accuracy drops significantly when the parts are occluded. This hugely limits the usability of applications that involve interaction with external objects, such as sport training or exercising systems. The problem becomes more critical when Kinect incorrectly perceives body parts. This is because applications have limited information about the recognition correctness, and using those parts to synthesize body postures would result in serious visual artifacts. In this paper, we propose a new method to reconstruct valid movement from incomplete and noisy postures captured by Kinect. We first design a set of measurements that objectively evaluates the degree of reliability on each tracked body part. By incorporating the reliability estimation into a motion database query during run time, we obtain a set of similar postures that are kinematically valid. These postures are used to construct a latent space, which is known as the natural posture space in our system, with local principle component analysis. We finally apply frame-based optimization in the space to synthesize a new posture that closely resembles the true user posture while satisfying kinematic constraints. Experimental results show that our method can significantly improve the quality of the recognized posture under severely occluded environments, such as a person exercising with a basketball or moving in a small room.
Background:
Several studies indicate that professional athletes can successfully return to competition after surgical treatment of femoroacetabular impingement (FAI). However, little is known about ...sports and activity levels after FAI surgery in the general patient population.
Hypothesis/Purpose:
The purpose was to determine the sports behavior, satisfaction with sports ability, and activity levels in a consecutive cohort of patients with FAI who were treated by surgical hip dislocation. The hypothesis was that the majority of patients (>75%) would be active in sports at follow-up.
Study Design:
Case series; Level of evidence, 4.
Methods:
This retrospective study included 153 patients (mean age, 30.0 years; 40.5% female) with 192 hips treated. Sports behavior and satisfaction were determined at a mean follow-up of 59.4 months with the use of a questionnaire. Activity levels at follow-up were assessed by the Hip Sports Activity Scale (HSAS) and the University of California, Los Angeles (UCLA) activity scale.
Results:
Of 126 patients who were regularly active in sports before surgery, 107 (85%) were so at follow-up. Nineteen patients (12.4%) stopped participating in regular sports, and 8 (5.2%) commenced with sports after the operation. The most popular activities before surgery were skiing (22%), cycling (22%), jogging (20%), and soccer (13%). At follow-up, most patients were engaged in cycling (23%), fitness/weight training (20%), skiing (18%), and jogging (11%). Of all patients, 75% were satisfied with their sports ability, and 25% were not. Moreover, 60.3% stated that their sports ability had improved after surgery, 20.5% declared no change, and 19.2% were subjectively deteriorated. The mean pain level during sports was rated to be 2.1 according to the visual analog scale. The mean HSAS score was 3.5 (range, 0-8), and the mean UCLA score was 7.7 (range, 3-10); male patients reported significantly higher scores than did female patients on the HSAS (4.1 vs 2.7, respectively) and UCLA scale (8.2 vs 7.0, respectively).
Conclusion:
The vast majority of patients with FAI who are treated by surgical hip dislocation return to sports activities, and most patients are satisfied with their sports ability at midterm follow-up. Activity levels are significantly higher in male patients, but this does not yield higher satisfaction rates.
There is paucity of literature on dietary treatment in glycogen storage disease (GSD) type IV and formal guidelines are not available. Traditionally, liver transplantation was considered the only ...treatment option for GSD IV. In light of the success of dietary treatment for the other hepatic forms of GSD, we have initiated this observational study to assess the outcomes of medical diets, which limit the accumulation of glycogen. Clinical, dietary, laboratory, and imaging data for 15 GSD IV patients from three centres are presented. Medical diets may have the potential to delay or prevent liver transplantation, improve growth and normalize serum aminotransferases. Individual care plans aim to avoid both hyperglycaemia, hypoglycaemia and/or hyperketosis, to minimize glycogen accumulation and catabolism, respectively. Multidisciplinary monitoring includes balancing between traditional markers of metabolic control (ie, growth, liver size, serum aminotransferases, glucose homeostasis, lactate, and ketones), liver function (ie, synthesis, bile flow and detoxification of protein), and symptoms and signs of portal hypertension.
Accidental allergic reactions to food are frequent and can be severe and even fatal.
We sought to analyze the culprit food products and levels of unexpected allergens in accidental reactions.
A ...prospective cohort study was conducted in adults (n = 157) with a physician-confirmed diagnosis of food allergy. During a 1-year follow-up, 73 patients reported accidental allergic reactions and the culprit food products. Food samples received (n = 51) were analyzed for a wide range of suspected noningredient allergens, and risk was quantified.
A very diverse range of food products was responsible for the unexpected allergic reactions. Thirty-seven percent (19/51) of products analyzed had 1 to 4 culprit allergens identified that were not supposed to be present according to the ingredient declaration. Concentrations varied from 1 to 5000 mg of protein of the allergenic food per kilogram of food product and were greatest for peanut, milk, and sesame. Milk proteins posed the highest estimated risk for objective allergic reactions. The intake of culprit allergens by patients varied considerably. For those cases in which culprit allergens were detected, the intake of at least 1 allergen exceeded the reference dose or a culprit allergen with a yet unknown reference dose was present. Both patient neglect of precautionary allergen labeling statements and omission of using a precautionary allergen labeling statement by food manufacturers seem to contribute to accidental reactions.
A wide range of food products are causing accidental reactions in patients with food allergy. Eight different allergens not declared on the ingredient lists were detected in the culprit food products, all of which were representative of allergens regulated in the European Union.
Display omitted
The rising prevalence of antimicrobial resistance in
serovars Typhi and Paratyphi A, causative agents of typhoid and paratyphoid, have led to fears of untreatable infections. Of specific concern is ...the emerging resistance against azithromycin, the only remaining oral drug to treat extensively drug resistant (XDR) typhoid. Since the first report of azithromycin resistance from Bangladesh in 2019, cases have been reported from Nepal, India, and Pakistan. The genetic basis of this resistance is a single point mutation in the efflux pump AcrB (R717Q/L). Here, we report 38 additional cases of azithromycin-resistant (AzmR)
Typhi and Paratyphi A isolated in Bangladesh between 2016 and 2018. Using genomic analysis of 56 AzmR isolates from South Asia with AcrB-R717Q/L, we confirm that this mutation has spontaneously emerged in different
Typhi and Paratyphi A genotypes. The largest cluster of AzmR Typhi belonged to genotype 4.3.1.1; Bayesian analysis predicts the mutation to have emerged sometime in 2010. A travel-related Typhi isolate with AcrB-R717Q belonging to 4.3.1.1 was isolated in the United Kingdom, increasing fears of global spread. For real-time detection of AcrB-R717Q/L, we developed an extraction-free, rapid, and low-cost mismatch amplification mutation assay (MAMA). Validation of MAMA using 113 AzmR and non-AzmR isolates yielded >98% specificity and sensitivity versus phenotypic and whole-genome sequencing assays currently used for azithromycin resistance detection. With increasing azithromycin use, AcrB-R717Q/L is likely to be acquired by XDR strains. The proposed tool for active detection and surveillance of this mutation may detect pan-oral drug resistance early, giving us a window to intervene.
In the early 1900s, with mortality of ∼30%, typhoid and paratyphoid ravaged parts of the world; with improved water, sanitation, and hygiene in resource-rich countries and the advent of antimicrobials, mortality dwindled to <1%. Today, the burden rests disproportionately on South Asia, where the primary means for combatting the disease is antimicrobials. However, prevalence of antimicrobial resistance is rising and, in 2016, an extensively drug resistant Typhi strain triggered an ongoing outbreak in Pakistan, leaving only one oral drug, azithromycin, to treat it. Since the description of emergence of azithromycin resistance, conferred by a point mutation in
(AcrB-R717Q/L) in 2019, there have been increasing numbers of reports. Using genomics and Bayesian analysis, we illustrate that this mutation emerged in approximately 2010 and has spontaneously arisen multiple times. Emergence of pan-oral drug resistant
Typhi is imminent. We developed a low-cost, rapid PCR tool to facilitate real-time detection and prevention policies.
Content-based image retrieval (CBIR) has attracted much attention during the past decades for its potential practical applications to image database management. A variety of relevance feedback (RF) ...schemes have been designed to bridge the gap between low-level visual features and high-level semantic concepts for an image retrieval task. In the process of RF, it would be impractical or too expensive to provide explicit class label information for each image. Instead, similar or dissimilar pairwise constraints between two images can be acquired more easily. However, most of the conventional RF approaches can only deal with training images with explicit class label information. In this paper, we propose a novel discriminative semantic subspace analysis (DSSA) method, which can directly learn a semantic subspace from similar and dissimilar pairwise constraints without using any explicit class label information. In particular, DSSA can effectively integrate the local geometry of labeled similar images, the discriminative information between labeled similar and dissimilar images, and the local geometry of labeled and unlabeled images together to learn a reliable subspace. Compared with the popular distance metric analysis approaches, our method can also learn a distance metric but perform more effectively when dealing with high-dimensional images. Extensive experiments on both the synthetic data sets and a real-world image database demonstrate the effectiveness of the proposed scheme in improving the performance of the CBIR.