Because of high dimensionality, correlation among covariates, and noise contained in data, dimension reduction (DR) techniques are often employed to the application of machine learning algorithms. Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), and their kernel variants (KPCA, KLDA) are among the most popular DR methods. Recently, Supervised Kernel Principal Component Analysis (SKPCA) has been shown as another successful alternative. In this paper, brief reviews of these popular techniques are presented first. We then conduct a comparative performance study based on three simulated datasets, after which the performance of the techniques are evaluated through application to a pattern recognition problem in face image analysis. The gender classification problem is considered on MORPH-II and FG-NET, two popular longitudinal face aging databases. Several feature extraction methods are used, including biologically-inspired features (BIF), local binary patterns (LBP), histogram of oriented gradients (HOG), and the Active Appearance Model (AAM). After applications of DR methods, a linear support vector machine (SVM) is deployed with gender classification accuracy rates exceeding 95% on MORPH-II, competitive with benchmark results. A parallel computational approach is also proposed, attaining faster processing speeds and similar recognition rates on MORPH-II. Our computational approach can be applied to practical gender classification systems and generalized to other face analysis tasks, such as race classification and age prediction.

This paper focuses on YOLO-LITE, a real-time object detection model developed to run on portable devices such as a laptop or cellphone lacking a Graphics Processing Unit (GPU). The model was first trained on the PASCAL VOC dataset then on the COCO dataset, achieving a mAP of 33.81% and 12.26% respectively. YOLO-LITE runs at about 21 FPS on a non-GPU computer and 10 FPS after implemented onto a website with only 7 layers and 482 million FLOPS. This speed is 3.8 × faster than the fastest state of art model, SSD MobilenetvI. Based on the original object detection algorithm YOLOV2, YOLO-LITE was designed to create a smaller, faster, and more efficient model increasing the accessibility of real-time object detection to a variety of devices.

为解决修水县城区供水高峰期的供需矛盾和马家坳水厂的备用问题,保障城市供水安全,项目采用浸没式超滤膜技术对罗桥水厂实施改造.结果表明:浸没式超滤膜系统低液位条件下虹吸出水,实现了短流程水质稳定产水;在原水浊度波动较大的条件下,膜出水水质稳定,出水浊度低于0.04 NTU;在处理水量超负荷50％的情况下,系统运行稳定,跨膜压差为10 kPa;运行费用较低.项目的实施为浸没式超滤膜系统应用于老水厂升级改造,提质增效,挖掘潜能积累了经验.

We perform preliminary studies on a large longitudinal face database MORPH-II, which is a benchmark dataset in the field of computer vision and pattern recognition. First, we summarize the inconsistencies in the dataset and introduce the steps and strategy taken for cleaning. The potential implications of these inconsistencies on prior research are introduced. Next, we propose a new automatic subsetting scheme for evaluation protocol. It is intended to overcome the unbalanced racial and gender distributions of MORPH-II, while ensuring independence between training and testing sets. Finally, we contribute a novel global framework for age estimation that utilizes posterior probabilities from the race classification step to compute a racecomposite age estimate. Preliminary experimental results on MORPH-II are presented.

This paper outlines the steps taken toward pre-processing the 55,134 images of the MORPH-II non-commercial dataset. Following the introduction, section two begins with an overview of each step in the pre-processing pipeline. Section three expands upon each stage of the process and includes details on all calculations made, by providing the OpenCV functionality paired with each step. The last portion of this paper discusses the potential improvements to this pre-processing pipeline that became apparent in retrospect.

In this project, competition-winning deep neural networks with pretrained weights are used for image-based gender recognition and age estimation. Transfer learning is explored using both VGG19 and VGGFace pretrained models by testing the effects of changes in various design schemes and training parameters in order to improve prediction accuracy. Training techniques such as input standardization, data augmentation, and label distribution age encoding are compared. Finally, a hierarchy of deep CNNs is tested that first classifies subjects by gender, and then uses separate male and female age models to predict age. A gender recognition accuracy of 98.7% and an MAE of 4.1 years is achieved. This paper shows that, with proper training techniques, good results can be obtained by retasking existing convolutional filters towards a new purpose.

Benchmark dose estimation is widely used in various regulatory and industrial settings to estimate acceptable exposure levels to hazardous or toxic agents by predefining a level of excess risk (US EPA in Benchmark dose technical guidance document. Technical Report #EPA/100/R-12/001. U.S. Environmental Protection Agency, Washington, DC, 2012). Although benchmark dose estimation is a popular method for identifying exposure levels of agents, there are some limitations and cautions on use of this methodology. One such concern is choice of the underlying risk model. Recently, advances have been made using Bayesian model averaging to improve benchmark dose estimation in the face of model uncertainty. Herein we employ the strategies of Bayesian model averaging to build model averaged estimates for the benchmark dose. The methodology is demonstrated via a simulation study and with real data.

ABSTRACT In this study, propanedioic acid was investigated as a potential additive in poly(phthalazine ether sulfone ketone) (PPESK)/ N ‐methyl‐2‐pyrrolidone solutions. Compared with poly(ethylene glycol) with a molecular weight of 1000 and Tween 80 as additives, different phenomena were observed: (1) both fingerlike and spongelike structures of asymmetric ultrafiltration membranes were induced by rapid gelation, and (2) a spongelike structure membrane with a high pure water flux was obtained under a high gelation rate. Moreover, the PPESK membrane formation process was recorded with a recently developed optical microscopy (OM)–charge‐coupled device (CCD) camera experimental system. The predicted membrane structure with an OM–CCD system gave good correspondence with the final membrane structure and performance, as detected by scanning electron microscopy. © 2014 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132 , 41621.

Poly(phthalazinone-ether-sulfone) (PPES) polymer is a relatively newly developed material with a bis(4-fluorodiphenyl) sulfone group. The formation of the PPES membrane by wet-phase inversion can proceed according to a slow or fast gelation method. These formation mechanisms were studied experimentally. The resulting membrane morphology was investigated using both optical and scanning electron micrography. The effects of PPES concentration and two additives, polyvinylpyrrolidone (PVP) and oxalic acid (OA), on the apparent viscosity and gelation rate of PPESK/NMP solutions and membrane performance have also been investigated. It was found that the gelation rate is important to obtain a sponge-like membrane structure, however favored by a fast gelation rate. The membrane obtained by a fast gelation rate showed a high pure water flux and rejection of bovine serum albumin (BSA), contrary to previous findings. On the basis of the experimental results, the actual membrane structure and pure water flux were related, and in agreement with the optical micrograph and gelation rate, respectively. The current results provide a fundamental insight in this novel copolymer, useful in future applications, especially in the membrane formation process.

Hydrophilic modification of high-density polyethylene membrane was performed with the introduction of peroxide onto the membrane surface by corona discharge treatment, followed by graft polymerization with acrylic acid. The surface graft polymerization was confirmed by attenuated total reflection-infrared and X-ray photo spectroscopy. Surface hydrophilicity was investigated by measuring the contact angles and moisture adsorption of the modified membrane. The results indicated that the modified membrane showed a significant increase in hydrophilicity. And water flux of the film increased from 102 L/m(2) h of the virgin film to 200 L/m(2) h of the grafted film, while the bovine serum albumin solution flux of the membrane increased from 43 L/m(2) h of the virgin film to 98 L/m(2) h of the modified membrane.

ADVERTISEMENT RETURN TO ISSUEPREVAddition/CorrectionCorrection to Preparation and Characterization of Poly(phthalazinone ether sulfone) Hollow Fiber Ultrafiltration MembranesYanbin Yun*, Panyue Zhang*, Minghang Zhu, Chao Liu, Lihua Wang, Cuixian Chen, and Jiding LiCite this: Langmuir 2012, 28, 28, 10627Publication Date (Web):March 15, 2012Publication History Received22 June 2011Revised13 March 2012Published online2 April 2012Published inissue 17 July 2012https://pubs.acs.org/doi/10.1021/la300239chttps://doi.org/10.1021/la300239ccorrectionACS PublicationsCopyright © 2012 American Chemical Society. This publication is available under these Terms of Use. Request reuse permissions This publication is free to access through this site. Learn MoreArticle Views742Altmetric-Citations2LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InRedditEmail PDF (109 KB) Get e-AlertscloseSupporting Info (1)»Supporting Information Supporting Information Get e-Alerts

Bilayer polyimide film with superparamagnetic response and conductive surface has been exploited as an important candidate for electromagnetic interference shielding material. Poly-imide matrix was derived from 4,4’-oxydianiline and 3,3’,4,4’-benzophenonetetracarboxylic dianhydride. Iron (III) 2,4-pentanedionate (Fe(acac)3) and (1,1,1-trifluoro-2,4-pentadionato) silver(I) (AgTFA) were chosen as magnetic nanoparticles precursor and silver source, respectively. At magnetic polyimide layer, in situ method allowed the Fe(acac) 3 to decompose to γ-Fe2O3, exhibiting typical superparamagnetic response. The conductive surface-silvered poly-imide layer was prepared via in situ single-stage self-metallization technique by thermal curing of the AgTFA-contained poly(amic acid). The reduction of silver(I) and further aggregation of silver atoms gave the conductive polyimide surface. The surface square resistance for the bilayer film of 0.1 Ω/square could be obtained. The structure and the properties of final bilayer f...

The synthesis of quarternized ammonium cationic polyvinyl alcohol (QPVA(+)) is studied in this article. The effect of synthesis conditions on the degree of substitution (DS) is studied to acquire the optimal synthesis conditions. Polyelectrolyte and polyelectrolyte complex (PEC) membranes of QPVA(+) and phosphatized anionic polyvinyl alcohol (PPVA(-)) are made by solution casting method. It is found that the flux of polyelectrolyte and PEC membrane, consisting of QPVA(+) and PPVA(-) (1 : 1 in molar ratio), is 403 g/m(2) h, and its separation factor is 970 when the membrane is used in the pervaporation separation of ethanol/water. It is concluded that the hydrophilicity of the membrane and the rigidity of molecular chain segment are enhanced when the anionic and cationic groups are introduced in the PVA, and the membrane separation capability could be adjusted by changing polyelectrolyte material type and cation/anion mole ratio in the polyelectrolyte and PEC. (C) 2010 Wiley Periodicals, Inc. J Appl Polym Sci 119: 2584-2594, 2011

This paper reports an evaluation of the effect of both organic and inorganic additives on the gelation kinetics, morphology and separation properties of asymmetric poly(phthalazine ether ketone) (PPEK) hollow fiber ultrafiltration membranes. Using an online optical microscope-CCD camera for observation of precipitation front movement (X), a good linear relationship between X2 and time was observed for the whole gelation process with the correlation coefficient (R2) higher than 0.98. It was found that the pure water flux of membranes with additives were lower than neat PPEK membranes. This observation might be attributed to the decrease in the face gelation velocity caused by the extra organic additives. The water flux of PEG600 was higher than that of Tween80 at the same additive concentration. When OA concentration increased from 3.5 to 5.0wt.%, the pure water flux increased from 332 to 367L/m2h and the rejection decreased from 98.6 to 96.8%. The PPEK hollow fiber membranes exhibited finger-like macro-voids structure and a skin layer at the outer-most region of the hollow fiber. Evaluation of the membrane chemical resistance with various chemical solutions indicated that PPEK membranes had good chemical resistance against HCl, H2SO4, HNO3, CH3COOH, OA, NaOH and H2O2, while the NaClO solution can greatly damage the membrane mechanical strength. Static sorption tests of PPEK membrane revealed that PPEK membrane adsorbed more bacitracin than BSA due to the smaller size of bacitracin.

In this work we aim to study rigorously the facial age estimation in a multiethnic environment with 39 possible combination of four feature normalization methods, two simple feature fusion methods, two feature selection methods, and three face representation methods as Gabor, AAM and LBP. First, Gabor feature is extracted as facial representation for age estimation. Inspired by [3], we further fuse the global Active Appearance Model (AAM) and the local Gabor features as the representation of faces. Combining with feature selection schemes such as Least Angle Regression (LAR) and sequential selection, an advanced age estimation system is proposed on the fused features. Systematic comparative of 39 experiments demonstrate that (1) As a single facial representation, Gabor features surprisedly outperform LBP features or even AAM features. (2) With global/local feature fusion scheme, fused Gabor and AAM or fused LBP and AAM features can achieve significant improvement in age estimation over single feature representation alone.