基于多特征空间的三维目标离面旋转识别

基于识别目标真假和空间位置的需要,将180幅训练图像分成了4个训练集,分别计算每个训练集的特征向量,分析结果表明每个训练集可以用3个特征向量来表示,这样用12个特征向量就可以建立目标的4个特征空间,使目标分解与重构过程大大简化。利用待识别目标向量与重构向量之间的关系,不仅可以判别目标的真假,还可以确定目标所在的空间位置。模拟结果表明,提出的目标多特征空间的建立方法和目标识别准则是有效的,能够实现离面旋转条件下三维目标真假和空间位置的识别。     

Development of an Optimized Feature Extraction Algorithm for Throat Signal Analysis

In this paper, we present a speech recognition system using a throat microphone. The use of this kind of microphone minimizes the impact of environmental noise. Due to the absence of high frequencies and the partial loss of formant frequencies, previous systems using throat microphones have shown a lower recognition rate than systems which use standard microphones. To develop a high performance automatic speech recognition (ASR) system using only a throat microphone, we propose two methods. First, based on Korean phonological feature theory and a detailed throat signal analysis, we show that it is possible to develop an ASR system using only a throat microphone, and propose conditions of the feature extraction algorithm. Second, we optimize the zero‐crossing with peak amplitude (ZCPA) algorithm to guarantee the high performance of the ASR system using only a throat microphone. For ZCPA optimization, we propose an intensification of the formant frequencies and a selection of cochlear filters. Experimental results show that this system yields a performance improvement of about 4% and a reduction in time complexity of 25% when compared to the performance of a standard ZCPA algorithm on throat microphone signals.     

大规模数据的L1惩罚分位数回归方法研究--基于特征筛选和随机抽样方法

为解决大规模数据在进行回归分析时存在的计算内存不足和运行时间较长的问题,提出两个新的回归分析方法:先筛选后抽样的大规模数据L₁惩罚分位数回归方法(FSSLQR)和先抽样后筛选的大规模数据L₁惩罚分位数回归方法(SFSLQR),其数值模拟和实际应用结果表明:FSSLQR和SFSLQR方法不仅能够显著降低计算内存和运行时间,而且其估计预测和变量选择的结果与全量L₁惩罚分位数回归基本一致。此外,与Xu等(2018)提出的大规模数据的L1惩罚分位数回归方法(SLQR)相比,FSSLQR和SFSLQR方法在估计预测、变量选择和运行时间等方面都更具优势。     

Bird Species Identification Using Spectrogram Based on Multi-Channel Fusion of DCNNs

Deep convolutional neural networks (DCNNs) have achieved breakthrough performance on bird species identification using a spectrogram of bird vocalization. Aiming at the imbalance of the bird vocalization dataset, a single feature identification model (SFIM) with residual blocks and modified, weighted, cross-entropy function was proposed. To further improve the identification accuracy, two multi-channel fusion methods were built with three SFIMs. One of these fused the outputs of the feature extraction parts of three SFIMs (feature fusion mode), the other fused the outputs of the classifiers of three SFIMs (result fusion mode). The SFIMs were trained with three different kinds of spectrograms, which were calculated through short-time Fourier transform, mel-frequency cepstrum transform and chirplet transform, respectively. To overcome the shortage of the huge number of trainable model parameters, transfer learning was used in the multi-channel models. Using our own vocalization dataset as a sample set, it is found that the result fusion mode model outperforms the other proposed models, the best mean average precision (MAP) reaches 0.914. Choosing three durations of spectrograms, 100 ms, 300 ms and 500 ms for comparison, the results reveal that the 300 ms duration is the best for our own dataset. The duration is suggested to be determined based on the duration distribution of bird syllables. As for the performance with the training dataset of BirdCLEF2019, the highest classification mean average precision (cmAP) reached 0.135, which means the proposed model has certain generalization ability.     

基于渐进式生成对抗网络的舰船红外图像仿真

红外图像仿真在红外导引头设计、仿真训练中起到十分关键的作用。针对如何生成高分辨率、视觉特征可控的红外图像,提出了一种基于渐进式生成对抗网络的红外图像仿真方法。本文利用舰船模型的红外图像数据集训练了图像合成网络,输入随机特征向量,输出高分辨率的红外仿真图像;设计了图像编码网络,实现红外图像到特征向量的转换;利用Logistic回归方法,在特征向量域找到了控制红外图像角度特征的方向向量,并据此生成了不同角度的舰船模型仿真图像;最后通过均值哈希算法和平均结构相似性算法来定量评价仿真图像和真实图像的差异,实验结果表明仿真的红外图像和真实图像的相似度很高,可以为真实舰船的可控化红外图像仿真提供参考。     

两种陆生植物浸提液对蛋白核小球藻(Chlorella pyrenoidosa)的化感作用

通过合果芋(Syngonium podophyllum)和榕树(Ficus microcarpa)两种陆生植物不同浓度、不同部位浸提液对蛋白核小球藻化感作用的研究,结果表明:两种植物对蛋白核小球藻都具有化感作用但强度不同,榕树的抑制效果比合果芋强且稳定.从化感作用的效果上看,合果芋浸提液对蛋白核小球藻呈现先抑制后促进的作用特性,而榕树浸提液始终具有抑制作用.除榕树气生根浸提液外,在抑制作用期间内不同浓度浸提液对蛋白核小球藻的化感作用强度大致随浸提液浓度的增大而增强.不同陆生植物或同株植物不同部位浸提液的化感作用强度也有所不同,一般榕树叶(EC50,96h为13.16g·L-1)合果芋叶(EC50,96h为14.37g·L-1)榕树气生根(EC50,96h为30.66g·L-1).     

Inhibition of Interleukin-6-Induced Matrix Metalloproteinase-2 Expression and Invasive Ability of Lemon Peel Polyphenol Extract in Human Primary Colon Cancer Cells

Among matrix metalloproteinases (MMPs), MMP-9/2 are key enzymes involved in the proteolysis of extracellular matrices in the inflammatory process and in cancer. Since MMP-9/2 expression levels, activity, and secretion is up-regulated during inflammation in response to pro-inflammatory cytokines, such as interleukin-6 (IL-6), many efforts have been devoted to identifying factors that could inhibit the IL-6-induced MMP-9/2 expression. Up to now, several reports indicated that polyphenols from fruits and vegetables are among the major components of health promotion for their antioxidant properties and also for their anti-inflammatory and anti-cancer agents. Among plant derived polyphenols, lemon (Citrus limon) peel extract (LPE) shows anti-cancer properties in various cancer types. In our previous work, we demonstrated that LPE can reduce IL-6-induced migration/invasiveness and MMP-9/2 up-regulation in some gastric cancer cell lines. This study aims to exploit the anti-cancer properties of LPE using an in vitro system model of inflammation, consisting of IL-6-exposed human primary colon cancer cells. We first analyzed the effect of LPE on IL-6-induced cell migration and invasiveness by wound healing and Boyden chamber assay, respectively. The MMP-2 mRNA expression levels and gelatinolytic activity in the cell culture media were determined by q-PCR analysis and gelatin zymography, respectively, and finally, the effects of LPE on IL-6-induced JAK2/STAT3 signaling pathways have been investigated by Western blotting analysis. Our results show that LPE is able to inhibit the IL-6-dependent cell migration and invasiveness associated with the up-regulation of MMP-2 expression levels and that these effects are correlated to the STAT3 phosphorylation in human primary T88 and T93 colon cancer cells.     

Aroma Properties of Cocoa Fruit Pulp from Different Origins

Cocoa pulp occurs as a by-product of cocoa bean production and can be repurposed to different food applications, such as jams, fruit preparations and beverages, improving the sustainability of cocoa production, as well as the livelihoods of cocoa farmers. In this work, aroma-active compounds of fresh cocoa fruit pulps from different origins were investigated by applying aroma extract dilution analyses in combination with gas chromatography-mass spectrometry/olfactometry for identification. In total, 65 aroma-active compounds were determined in four different pulps originating from Indonesia, Vietnam, Cameroon, and Nicaragua. Vietnamese pulp showed the highest number of aroma-active regions, while Cameroonian pulp accounted for the lowest. Moreover, Cameroonian cocoa pulp showed the lowest FD factors. Overall, the odorants with the highest FD factors were trans-4,5-epoxy-(E)-decenal, 2- and 3-methylbutanoic acid, 3-(methylthio)propanal, 2-isobutyl-3-methoxypyrazine, (E,E)-2,4-nonadienal, (E,E)-2,4-decadienal, 4-vinyl-2-methoxyphenol, δ-decalactone, 3-hydroxy-4,5-dimethylfuran-2(5H)-one, dodecanoic acid, and linalool. This study provides insights into the aroma composition of fresh cocoa pulp from different origins for future food applications.     

Bioactive Pectin-Murta (Ugni molinae T.) Seed Extract Films Reinforced with Chitin Fibers

This study investigated the biocomposite pectin films enriched with murta (Ugni molinae T.) seed polyphenolic extract and reinforced by chitin nanofiber. The structural, morphological, mechanical, barrier, colorimetric, and antioxidant activity of films were evaluated. The obtained data clearly demonstrated that the addition of murta seed extract and the high load of chitin nanofibers (50%) provided more cohesive and dense morphology of films and improved the mechanical resistance and water vapor barrier in comparison to the control pectin film. The antioxidant activity ranged between 71% and 86%, depending on the film formulation and concentration of chitin nanofibers. The presented results highlight the potential use of chitin nanofibers and murta seed extract in the pectin matrix to be applied in functional food coatings and packaging, as a sustainable solution.     

Multi-Label Feature Selection Combining Three Types of Conditional Relevance

With the rapid growth of the Internet, the curse of dimensionality caused by massive multi-label data has attracted extensive attention. Feature selection plays an indispensable role in dimensionality reduction processing. Many researchers have focused on this subject based on information theory. Here, to evaluate feature relevance, a novel feature relevance term (FR) that employs three incremental information terms to comprehensively consider three key aspects (candidate features, selected features, and label correlations) is designed. A thorough examination of the three key aspects of FR outlined above is more favorable to capturing the optimal features. Moreover, we employ label-related feature redundancy as the label-related feature redundancy term (LR) to reduce unnecessary redundancy. Therefore, a designed multi-label feature selection method that integrates FR with LR is proposed, namely, Feature Selection combining three types of Conditional Relevance (TCRFS). Numerous experiments indicate that TCRFS outperforms the other 6 state-of-the-art multi-label approaches on 13 multi-label benchmark data sets from 4 domains.