自适应融合层级特征的混合退化图像复原算法

多种退化类型混合的图像比单一类型的退化图像降质更严重,很难建立精确模型对其复原,研究端到端的神经网络算法是复原的关键.现有的基于操作选择注意力网络的算法(operation-wiseattentionnetwork,OWAN)虽然有一定的性能提升,但是其网络过于复杂,运行较慢,复原图像缺乏高频细节,整体效果也有提升的空间.针对这些问题,提出一种基于层级特征融合的自适应复原算法.该算法直接融合不同感受野分支的特征,增强复原图像的结构;用注意力机制对不同层级的特征进行动态融合,增加模型的自适应性,降低了模型冗余;另外,结合L1损失和感知损失,增强了复原图像的视觉感知效果.在DIV2K,BSD500等数据集上的实验结果表明,该算法无论是在峰值信噪比和结构相似性上的定量分析,还是在主观视觉质量方面,均优于OWAN算法,充分证明了该算法的有效性.     

YOLSO: You Only Look Small Object

Small object detection is challenging and far from satisfactory. Most general object detectors suffer from two critical issues with small objects: (1) Feature extractor based on classification network cannot express the characteristics of small objects reasonably due to insufficient appearance information of targets and a large amount of background interference around them. (2) The detector requires a much higher location accuracy for small objects than for general objects. This paper proposes an effective and efficient small object detector YOLSO to address the above problems. For feature representation, we analyze the drawbacks in previous backbones and present a Half-Space Shortcut(HSSC) module to build a background-aware backbone. Furthermore, a coarse-to-fine Feature Pyramid Enhancement(FPE) module is introduced for layer-wise aggregation at a granular level to enhance the semantic discriminability. For loss function, we propose an exponential L1 loss to promote the convergence of regression, and a focal IOU loss to focus on prime samples with high classification confidence and high IOU. Both of them significantly improves the location accuracy of small objects. The proposed YOLSO sets state-of-the-art results on two typical small object datasets, MOCOD and VeDAI, at a speed of over 200 FPS. In the meantime, it also outperforms the baseline YOLOv3 by a wide margin on the common COCO dataset.     

面向分拣机器人的珍珠形状视觉检测方法

针对人工进行珍珠形状分拣效率低、精度不稳定等问题,提出基于机器视觉的珍珠形状检测方法;采用背光成像方式消除珍珠表面纹理和光泽的影响,对获取的珍珠图像进行同态滤波等预处理算法,提高图像对比度;为了解决相互接触珍珠影响珍珠轮廓提取的问题,采用分水岭算法对珍珠图像进行分割,得到了独立存在的珍珠个体,再通过连通域标记、质心算法对珍珠进行定位;根据国家标准对珍珠形状的规定,基于珍珠图像信息建立珍珠形状参数模型,对珍珠形状进行量化;实验结果表明,不同形状的珍珠样本的检测误差为0.63%,形状统计精度为100%,算法耗时24 ms;该方法可准确高效地对珍珠进行分拣分级,具有一定的实用价值。     

Subdivision based isogeometric analysis for geometric flows

We present a new isogeometric analysis (IGA) approach based on extended Loop subdivision scheme for solving various geometric flows defined on subdivision surfaces. The studied flows include the second-order, fourth-order, and sixth-order geometric flows, such as averaged mean curvature flow, constant mean curvature flow, and minimal mean-curvature-variation flow, which are generally derived by minimizing the associate energy functionals with $L^2$-gradient flow respectively. The geometric flows are discretized by means of subdivision based IGA, where the finite element space is formulated by the limit form of the extended Loop subdivision for different initial control meshes. The basis functions, consisting of quartic box-splines corresponding to each subdivided control mesh, are utilized to represent the geometry exactly. For the cases of the evolution of open surfaces with any shape boundary, high-order continuous boundary conditions derived from the mixed variational forms of the geometric flows should be implemented to be consistent with the isogeometric concept. For time discretization, we adopt an adaptive semi-implicit Euler scheme. By several numerical experiments, we study the convergence behaviors of the proposed approach for solving the geometric flows with high-order boundary conditions. Moreover, the numerical results also show the accuracy and efficiency of the proposed method.     

Hybrid hydrate processes for CO2/H2 mixture purification: A techno-economic analysis

In this work, hydrate based separation technique was combined with membrane separation and amine-absorption separation technologies to design hybrid processes for separation of CO₂/H₂ mixture. Hybrid processes are designed in the presence of different types of hydrate promoters. The conceptual processes have been developed using Aspen HYSYS. Proposed processes were simulated at different flow rates for the feed stream. A comprehensive cost model was developed for economic analysis of novel processes proposed in this study. Based on the results from process simulation and equipment sizing, the amount of total energy consumption, fixed cost, variable cost, and total cost were calculated per unit weight of captured CO₂ for various flow rates of feed stream and in the presence of different hydrate promoters. Results showed that combination of hydrate formation separation technique with membrane separation technology results in a CO₂ capture process with lowest energy consumption and total cost per unit weight of captured CO₂. As split fraction and heat of hydrate formation increases, the share of hydrate formation section in total energy consumption increases. When TBAB is applied as hydrate promoter, due to its higher hydrate separation efficiency, more amount of CO₂ is captured in hydrate formation section and consequently the total cost for process decreases considerably. Hybrid hydrate-membrane process in the presence of TBAB as hydrate promoter with 29.47 US$/ton CO₂ total cost is the best scheme for hybrid hydrate CO₂ capture process. Total cost for this process is lower than total cost for single MDEA-based absorption process as the mature technology for CO₂ capture.     

Use of Fourier-Transform Infrared Spectroscopy (FT-IR) for Monitoring Experimental Helicobacter pylori Infection and Related Inflammatory Response in Guinea Pig Model

Infections due to Gram-negative bacteria Helicobacter pylori may result in humans having gastritis, gastric or duodenal ulcer, and even gastric cancer. Investigation of quantitative changes of soluble biomarkers, correlating with H. pylori infection, is a promising tool for monitoring the course of infection and inflammatory response. The aim of this study was to determine, using an experimental model of H. pylori infection in guinea pigs, the specific characteristics of infrared spectra (IR) of sera from H. pylori infected (40) vs. uninfected (20) guinea pigs. The H. pylori status was confirmed by histological, molecular, and serological examination. The IR spectra were measured using a Fourier-transform (FT)-IR spectrometer Spectrum 400 (PerkinElmer) within the range of wavenumbers 3000–750 cm⁻¹ and converted to first derivative spectra. Ten wavenumbers correlated with H. pylori infection, based on the chi-square test, were selected for a K-nearest neighbors (k-NN) algorithm. The wavenumbers correlating with infection were identified in the W2 and W3 windows associated mainly with proteins and in the W4 window related to nucleic acids and hydrocarbons. The k-NN for detection of H. pylori infection has been developed based on chemometric data. Using this model, animals were classified as infected with H. pylori with 100% specificity and 97% sensitivity. To summarize, the IR spectroscopy and k-NN algorithm are useful for monitoring experimental H. pylori infection and related inflammatory response in guinea pig model and may be considered for application in humans.     

Evaluation of the localized corrosion on Mg electrode surface by electrochemical noise technique

The symmetrical and asymmetrical electrodes made of Mg were studied in 0.1-M NaCl electrolyte adjusted at pH 12. The statistical and wavelet methods were employed for analyzing the electrochemical current noise (ECN) signals. The asymmetric configuration was used for electrochemical detection of filiform corrosion on Mg electrode. The real time scale of the dominant transients of the asymmetric electrodes was detected on the basis of the maximum peak in the SDPS plots. The SDPS values of the real time scale crystals of the ECN signals resulting from asymmetrical electrodes increased with the increase in immersion time due to the onset of filiform corrosion.     

A novel adaptive sampling based methodology for feasible region identification of compute intensive models using artificial neural network

Identification of feasible region of operations in multivariate processes is a problem of interest in several fields. This is particularly challenging when the process model is black-box in nature and/or is computationally expensive, as analytical solutions are not available and the number of possible model evaluations is limited. An efficient methodology is required to identify samples where the model is evaluated for developing a computationally efficient surrogate model. In this work, an artificial neural network based surrogate model is proposed which is integrated with a statistical-based approach (Jack-knifing) to estimate the variance of the surrogate model prediction. This allows implementation of an adaptive sampling approach where new samples are identified close to the feasible region boundary or in regions of high prediction uncertainty. The proposed approach performs better than a previously published kriging based method for different dimensionality case studies.     

贵长猕猴桃皮HPLC指纹图谱及抗氧化谱效关系研究

目的:研究贵长猕猴桃皮HPLC指纹图谱与抗氧化活性的关系。方法:采用高效液相色谱法建立10批不同来源地贵长猕猴桃皮的指纹图谱,采用SPSS 13.0进行样品聚类分析和主成分分析;并通过DPPH和ABTS法测定其抗氧化活性,以EC50与共有峰峰面积数据为基础,应用皮尔逊相关分析法和逐步回归分析法研究谱效关系。结果:建立了10批贵长猕猴桃皮HPLC指纹图谱,确定了14个共有峰,相似度均>0.90。采用对照品比对方法指认了其中6个峰:2号峰没食子酸,5号峰原儿茶酸,9号峰表儿茶素,10号峰二氢槲皮素,11号峰4-香豆酸,12号峰落新妇苷,样本可聚为3类。皮尔逊相关分析与逐步回归分析结果显示:5号峰(原儿茶酸)和9号峰(表儿茶素)峰面积变化与贵长猕猴桃皮抗氧化活性呈显著正相关,对其抗氧化活性贡献度较大。结论:贵长猕猴桃皮具有较高的体外抗氧化活性,抗氧化活性的物质基础初步确定为贵长猕猴桃皮开发利用、质量评价提供依据。     

Feature selection methods for root-cause analysis among top-level product attributes

Manufacturing companies not only strive to deliver flawless products but also monitor product failures in the field to identify potential quality issues. When product failures occur, quality engineers must identify the root cause to improve any affected product and process. This root-cause analysis can be supported by feature selection methods that identify relevant product attributes, such as manufacturing dates with an increased number of product failures. In this paper, we present different methods for feature selection and evaluate their ability to identify relevant product attributes in a root-cause analysis. First, we compile a list of feature selection methods. Then, we summarize the properties of product attributes in warranty case data and discuss these properties regarding the challenges they pose for machine learning algorithms. Next, we simulate datasets of warranty cases, which emulate these product properties. Finally, we compare the feature selection methods based on these simulated datasets. In the end, the univariate filter information gain is determined to be a suitable method for a wide range of applications. The comparison based on simulated data provides a more general result than other publications, which only focus on a single use case. Due to the generic nature of the simulated datasets, the results can be applied to various root-cause analysis processes in different quality management applications and provide a guideline for readers who wish to explore machine learning methods for their analysis of quality data.