基于局部约束邻域嵌入的人脸画像照片合成

针对画像块和照片块在流形上的邻域关系并不能完全反映彼此内在数据结构的问题,提出一种基于局部约束邻域嵌入(LCNE)的画像-照片合成算法。首先,利用基于邻域嵌入(NE)的合成方法得到待合成照片或画像的初始估计;其次,根据待合成的照片块或画像块与训练集中的照片块或画像块的相似性来约束合成权值;然后,通过交替优化方法进行权值的确定和K-近邻的选择,并更新待合成目标块;最后,合并所有估计的照片块或画像块合成目标图像。与基于邻域嵌入的画像照片合成方法相比,所提方法合成图像的结构相似度提高0.0503,脸识别准确率提高14%。实验结果表明,该方法解决了基于NE方法导致的邻域之间兼容性不强的问题,能大大减少合成图像上的噪声及变形。     

A comparison of clustering-based privacy-preserving collaborative filtering schemes

Privacy-preserving collaborative filtering (PPCF) methods designate extremely beneficial filtering skills without deeply jeopardizing privacy. However, they mostly suffer from scalability, sparsity, and accuracy problems. First, applying privacy measures introduces additional costs making scalability worse. Second, due to randomness for preserving privacy, quality of predictions diminishes. Third, with increasing number of products, sparsity becomes an issue for both CF and PPCF schemes.In this study, we first propose a content-based profiling (CBP) of users to overcome sparsity issues while performing clustering because the very sparse nature of rating profiles sometimes do not allow strong discrimination. To cope with scalability and accuracy problems of PPCF schemes, we then show how to apply k-means clustering (KMC), fuzzy c-means method (FCM), and self-organizing map (SOM) clustering to CF schemes while preserving users’ confidentiality. After presenting an evaluation of clustering-based methods in terms of privacy and supplementary costs, we carry out real data-based experiments to compare the clustering algorithms within and against traditional CF and PPCF approaches in terms of accuracy. Our empirical outcomes demonstrate that FCM achieves the best low cost performance compared to other methods due to its approximation-based model. The results also show that our privacy-preserving methods are able to offer precise predictions.     

An electricity production study by Rhodobacter sphaeroides

Microbial fuel cells are a type of bio-electrochemical system which can capture electrons produced by microorganisms. In this study, it is aimed to increase the electrogenic capacity of photosynthetic microbial fuel cells in a newly designed fuel cell. Rhodobacter sphaeroides, non-sulfur purple bacterium, was grown in anodic part of the fuel cells against permanganate as a cathodic electrolyte. Platine and graphite were used as the anodic and cathodic electrode, consequently. The distance between two electrodes was 1.5 cm. The concentrations of cathodic electrolyte were optimized. By the end, it is made to reach the highest anode potential (1.006 V) and electrogenic capacity (851.82 μA) in 5 mM permanganate concentration. The internal resistance was calculated as 1170 Ω. In these conditions, the current density with 2.1 cm² cathodic surface is 405.63 mA/m². These are the highest value of electricity generation potential of R. sphaeroides within the known PMFCs.     

A magnified beam algorithm to determine the coupling ratios of R/C coupled shear wall

The structural behaviour of a coupled shear wall (CSW) is greatly affected by the geometrical and mechanical properties of the coupling beams. The coupling ratio (r) is defined to represent the proportion of overturning moment resisted by each individual shear wall. The coupling ratio (r) is considered to be a key parameter in the nonlinear analysis and design of CSWs. In this study, the coupling ratios are determined on the basis of magnified beam analysis, using a statistical approach. An adequate number of CSWs have been selected with varying geometrical and mechanical properties, and the coupling ratios have been determined for each case using a magnified beam algorithm. The structural analyses have been also performed by the finite element modelling, and relationships for the estimation of the coupling ratios are provided using nonlinear regression analyses. The magnified beam approach adapted in this study is therefore proven to be capable of providing accurate and practical estimates for the coupling ratios of R/C CSWs. The results are quite promising for further research of any expanded data sets. Copyright © 2009 John Wiley & Sons, Ltd.     

Investigation of cylinder pressure for H2/CH4 mixtures at different loads

The purpose of this study is to experimentally analyze the performance and the pollutant emissions of a four-stroke spark-ignition engine operating on natural gas–hydrogen blends of 0%, 10%, 20% and 30% at full load and 65% load for different excess air ratios. This present work was carried out on a Ford engine. This is a four-stroke cycle four-cylinder spark-ignition engine with a bore × stroke of 80.6 × 88 mm and a compression ratio of 10:1. Experiments were made at a constant engine speed of 2000 rpm. CO, CO₂ and HC emission values and cylinder pressures were measured. The results showed that while the excess air ratio increases, CO and CO₂ emission values decrease.     

Robust watermarking of fingerprint images

This paper introduces two spatial methods in order to embed watermark data into fingerprint images, without corrupting their features. The first method inserts watermark data after feature extraction, thus preventing watermarking of regions used for fingerprint classification. The method utilizes an image adaptive strength adjustment technique which results in watermarks with low visibility. The second method introduces a feature adaptive watermarking technique for fingerprints, thus applicable before feature extraction. For both of the methods, decoding does not require original fingerprint image. Unlike most of the published spatial watermarking methods, the proposed methods provide high decoding accuracy for fingerprint images. High data hiding and decoding performance for color images is also observed.     

Comparative evaluation of 3D vs. 2D modality for automatic detection of facial action units

Automatic detection of facial expressions attracts great attention due to its potential applications in human–computer interaction as well as in human facial behavior research. Most of the research has so far been performed in 2D. However, as the limitations of 2D data are understood, expression analysis research is being pursued in 3D face modality. 3D can capture true facial surface data and is less disturbed by illumination and head pose. At this junction we have conducted a comparative evaluation of 3D and 2D face modalities. We have investigated extensively 25 action units (AU) defined in the Facial Action Coding System. For fairness we map facial surface geometry into 2D and apply totally data-driven techniques in order to avoid biases due to design. We have demonstrated that overall 3D data performs better, especially for lower face AUs and that there is room for improvement by fusion of 2D and 3D modalities. Our study involves the determination of the best feature set from 2D and 3D modalities and of the most effective classifier, both from several alternatives. Our detailed analysis puts into evidence the merits and some shortcomings of 3D modality over 2D in classifying facial expressions from single images.     

Experimental investigation of performance and emissions of the SICAI-hybrid engine systems

The use of hybrid electrical engines can provide more efficiency by reducing fuel consumption and emissions. In the research, the experimental studies on the created hybrid electric engine were presented. The hybrid engine combines an electric motor with the internal combustion engine (ICE) which is operated under spark assisted controlled auto-ignition (SICAI) combustion mode with the alternative fuels consisting of different ratios of methane–hydrogen blends. In order to establish the hybrid engine, firstly, efficiency graphs of the electrical motor were obtained experimentally. The battery charge status was also checked. The operating range of the SICAI engine in the hybrid system was identified considering performance and efficiency parameters. Based on these parameters, a hybrid algorithm was established to control the operating of the created hybrid engine system. Thus, the experimental studies were carried out for 100% methane, 90% methane-10% hydrogen, 80% methane-20% hydrogen and, 70% methane-30% hydrogen blends (by volume) at wide opening throttle (WOT) and, 50% WOT positions. Consequently, the results were discussed in terms of efficiency and emissions.     

Neuroprotective potential of some terebinth coffee brands and the unprocessed fruits of Pistacia terebinthus L. and their fatty and essential oil analyses

In the current study, neuroprotective effects of the ethyl acetate and methanol extracts of four terebinth coffee brands and the fruits of Pistacia terebinthus L. were investigated through enzyme inhibition tests against acetylcholinesterase, butyrylcholinesterase, and tyrosinase as well as antioxidant test systems. Antioxidant activity was measured using radical scavenging activity tests and metal-related tests including metal-chelation capacity, ferric-reducing antioxidant power (FRAP), and phosphomolybdenum reducing power (PRAP). The fatty oils of the coffee brands and the fruits and the fruit essential oil were examined by GC–MS. Total phenol and flavonoid contents were calculated spectrophotometrically. The extracts had moderate inhibition against butyrylcholinesterase (9.78–45.74% at 200 μg mL⁻¹) and potent scavenging activity against DPPH. They exerted strong activity in FRAP and metal-chelation tests and modest activity in PRAP test. Oleic acid was identified as the major fatty acid in the fatty oils, while α-pinene (26.31%) was dominant in the essential oil.