Visual Speech Recognition with Lightweight Psychologically Motivated Gabor Features

Extraction of relevant lip features is of continuing interest in the visual speech domain. Using end-to-end feature extraction can produce good results, but at the cost of the results being difficult for humans to comprehend and relate to. We present a new, lightweight feature extraction approach, motivated by human-centric glimpse-based psychological research into facial barcodes, and demonstrate that these simple, easy to extract 3D geometric features (produced using Gabor-based image patches), can successfully be used for speech recognition with LSTM-based machine learning. This approach can successfully extract low dimensionality lip parameters with a minimum of processing. One key difference between using these Gabor-based features and using other features such as traditional DCT, or the current fashion for CNN features is that these are human-centric features that can be visualised and analysed by humans. This means that it is easier to explain and visualise the results. They can also be used for reliable speech recognition, as demonstrated using the Grid corpus. Results for overlapping speakers using our lightweight system gave a recognition rate of over 82%, which compares well to less explainable features in the literature.     

云制造平台加工能力分享调度优化研究

云制造平台加工能力分享的调度优化是云制造模式的核心运营问题。为提高云制造平台对参与能力分享企业资源的利用率,解决实际云制造需求环境下企业级制造资源调度问题,达到云制造平台动态实时的要求,以最小化工期和成本为目标函数,构建了基于平台并面向分享的云加工能力调度模型。为快速求解多个云任务订单的最优调度序列,引入任务简易编码和随机权重,设计了改进的二阶粒子群算法,并提出可分享加工能力时间窗的更新策略。最后,通过算例验证云制造平台加工能力调度机制的可行性和有效性。     

Full physicochemical characterization of malic acid: Emphasis in the potential as food ingredient and application in pectin gels

Malic acid, a carboxylic acid most found in fruits, is a smooth taste substance used as flavoring and preservative agent in foods, although not as used as citric acid. There are no studies focusing in quantitative results or investigations on its physicochemical properties, useful to the food industry, or even the confirmation of its calcium chelating, buffer texturizer and antioxidant alleged properties. Thus, the aim of this work was the assessment of most physicochemical properties of malic acid, solid and in solution, that could be useful to the food industry understand its real potential. The following analyses were carried out: melting point; structure (NMR, XRD, FTIR and SEM/EDS); TGA/DTG; solubility, hygroscopicity; antioxidant activity, iron chelating and antibacterial activities and stability of pectin gels. The melting temperature found was 129.71 °C. TGA/DTG exhibited first loss of mass around 140 °C. In the temperature range of 10 to 55 °C, it exhibited a high solubility in water, from 48.12 to 61.49 (100w), respectively. The tested bacteria, related to food spoilage, were inhibited by DL-malic acid 10% or higher. Chelating and antioxidant activities showed expressive results even in 1% solution. Pectin gels with malic acid had stronger structure and less syneresis than citric acid gels. In addition, calcium chelating, buffer texturizer and antioxidant properties were confirmed. Thus, malic acid has potential to be applied in a wide variety of food products as fortified beverages, frozen and refrigerated items, oils, pectin gels, hard and soft candies, and biofilms, due to all the characteristics quantified.     

Serum pharmacochemistry combined with multiple data processing approach to screen the bioactive components and their metabolites in Mutan Cortex by ultra‐performance liquid chromatography tandem mass spectrometry

Root cortex of Paeonia suffruticosa Andrews (Paeoniaceae), known as Moutan Cortex (MC), is known to have anti‐allergic and anti‐inflammatory properties. However, the constituents absorbed into blood after oral administration of MC remain unknown. A sensitive and rapid method by ultra‐high‐pressure liquid chromatography–electrospray ionization–quadrupole‐time‐of‐flight mass spectrometry (UPLC‐ESI‐Q‐TOF‐MS) technology and the MetaboLynx^TM software combined with multiple data processing approach (Mdpa) was established to investigate the absorbed constituents in rats after oral administration of MC, providing unique high‐throughput capabilities for drug metabolism study. A hyphenated electrospray ionization and quadrupole‐time‐of‐flight analyzer was used for the determination of accurate mass of the fragment ion in negative mode, with excellent MS mass accuracy and enhanced data acquisition. This rapid automated analysis method was successfully applied for screening and identification of the constituents absorbed and metabolized studies of MC after oral administration to rats. A total of 46 peaks were obtained from MC, 41 of which were tentatively characterized. In the VIP‐plot of orthogonal partial least‐squares discriminant analysis, 23 interesting ions in serum samples were extracted, and 16 parent components and seven metabolites were detected in vivo. The integrative serum pharmacochemistry technique, UPLC‐ESI‐Q‐TOF‐MS, and Mdpa method were successfully applied for rapid discovery of multiple components from MC. Copyright © 2013 John Wiley & Sons, Ltd.     

一种新的磁共振图像处理流水线的设计与实现

磁共振图像的重建、后处理及可视化是磁共振成像（MRI）系统的重要组成部分.本文开发了一个新的用于磁共振图像重建、后处理及可视化的开源框架YAP（Yet Another Pipeline），利用此框架可以方便地构建图像处理流水线.与现有的一些其他开源框架相比，本文开发的框架具有如下特点：（1）采用基于接口的设计，可使用基于接口的插件对流水线的功能进行扩展；（2）允许用户使用编写脚本的方式构建图像处理流水线，编辑与修改流水线都很方便；（3）支持带有分支结构的流水线，便于流水线的构建与调试.目前，该框架已经在商用系统中获得了应用.     

A new class of conformable spectral observers for signal reconstruction

In this work, the design of spectral observers for signal reconstruction based on Kalman filters is performed and evaluated. The conformable derivative and the beta‐derivative were used to design the Kalman filters. Both derivatives satisfy the same formulas of the classical derivation, eg, the chain rule. The derivative order, the Ricatti equation parameters, and the observers tuning parameters were optimized using an optimization algorithm based on the bat's echolocation behavior (Bat algorithm). The simulation results showed the advantages of using the proposed observers for the signal reconstruction.     

Structural and luminescence characteristics of CuAlxIn1−xSe2 (0 < x ≤ 0.30) chalcopyrite solid solutions

1. Download : Download high-res image (112KB)
2. Download : Download full-size image

     

Effect of ultrasound on the synthesis of low-modulus zeolites from a metakaolin

It was studied the effect of ultrasonic processing (22 kHz) of the aqueous suspension of metakaolin, sodium hydroxide and alumina with a molar ratio 2Al₂Si₂O₇:12NaOH:2Al₂O₃ on the low-modulus zeolite synthesis processes. To investigate the XRD, SEM, IR, EDS had been used. It was shown that after ultrasonic processing, sodium aluminates are formed, what leads to a change in process of further synthesis. It was found that without ultrasonic processing on the stage of thermal treatment at 650 °C, SOD zeolite (|Na₆|[Al₆Si₆O₂₄]) and sodium aluminosilicate (Na₆Al₄Si₄O₁₇) are synthesized. In the sample after ultrasound during thermal treatment, only sodium aluminosilicates of cubic syngony (Na₆Al₄Si₄O₁₇ and Na₈Al₄Si₄O₁₈) are formed. It was demonstrated that sodium aluminosilicates are precursors for the formation of LTA zeolite (|Na₁₂|[Al₁₂Si₁₂O₄₈]). As a result zeolitization of sodium aluminosilicates after the hydrothermal crystallization in alkaline solution, the sonicated sample contained 97 wt% LTA. Without ultrasonic processing, the product of synthesis contained 50 wt% SOD and 40 wt% LTA.     

Evaluation of yarn defects by image processing technique

Yarn defect evaluation, since they can drastically deteriorate the quality of the product, is one of the most important problems in textile industry. In this work, the yarn defects are evaluated by image processing technique. Initially, images were taken by scanner, later the obtained images were modified by using several filters. Then, the yarn defects were identified base on their geometric shape and surface area. As the results show, image processing methods is excellently reliable in evaluating number and type of yarn defects.