Enhance the Performance of Directional Feature-based Palmprint Recognition by Directional Response Stability Measurement

Palmprint recognition is an emerging biometrics technology that has attracted increasing attention in recent years. Many palmprint recognition methods have been proposed, including traditional methods and deep learning-based methods. Among the traditional methods, the methods based on directional features are mainstream because they have high recognition rates and are robust to illumination changes and small noises. However, to date, in these methods, the stability of the palmprint directional response has not been deeply studied. In this paper, we analyse the problem of directional response instability in palmprint recognition methods based on directional feature. We then propose a novel palmprint directional response stability measurement (DRSM) to judge the stability of the directional feature of each pixel. After filtering the palmprint image with the filter bank, we design DRSM according to the relationship between the maximum response value and other response values for each pixel. Using DRSM, we can judge those pixels with unstable directional response and use a specially designed encoding mode related to a specific method. We insert the DRSM mechanism into seven classical methods based on directional feature, and conduct many experiments on six public palmprint databases. The experimental results show that the DRSM mechanism can effectively improve the performance of these methods. In the field of palmprint recognition, this work is the first in-depth study on the stability of the palmprint directional response, so this paper has strong reference value for research on palmprint recognition methods based on directional features.

Kernel Model Applied in Kernel Direct Discriminant Analysis for the Recognition of Face with Nonlinear Variations

一个基于核的判别式分析方法把核称为直接判别式分析被采用，它把直接线性的判别式分析的优点与核诡计的相结合。为了表明它的更好的坚韧性到真实的脸的复杂、非线性的变化，想象，例如照明，面部表情，规模和姿势变化，实验在 Olivetti 研究实验室，耶鲁和自我造的脸上被执行数据库。Theresults 与核主管部件分析和核相对照显示那线性判别式分析，方法能完成更低(7%) 用仅仅特征的一个很小的集合的错误率。而且，一个新改正的内核模型被建议改进识别性能。试验性的结果证实它的优势(1% 以识别评价) 到另外的多项式，核当模特儿。

F-Law collision and system state recognition

Using function one direction S-rough sets （function one direction singular rough sets）, f-law and F- law and the concept of law distance and the concept of system law collided by F-law are given. Using these concepts, state characteristic presented by system law collided by F-law and recognition of these states characteristic and recognition criterion and applications are given. Function one direction S-rough sets is one of basic forms of function S-rough sets （function singular rough sets）. Function one direction S-rough sets is importance theory and is a method in studying system law collision.

F-generation law and recognition of system law

If a system is not disturbed （or invaded） by some law, there is no doubt that each system will move according to the expected law and keep stable. Although such a fact often appears, some unknown law breaks into the system and leads it into turbulence. Using function one direction S-rough sets, this article gives the concept of the F-generation law in the system, the generation model of the F-generation law and the recognition method of the system law. Function one direction singular rough sets is a new theory and method in recognizing the disturbance law existing in the system and recognizing the system law.

■-law collision and system state recognition

Using dual function one direction S-rough sets, this article gives the f-law, the F-law, law distance and the concept of system law collided by the F-law. The characteristics presented by the system law collided by the F-law, the recognition of these characteristics and recognition criterion are also proposed. The dual function one direction S-rough sets is one of the basic forms of function S-rough sets. Its basic theory and application in the study of system law collision are reviewed.

F-interference law generation and its feature recognition

By using two-directional S-rough sets, the concepts of （f, f）-interference generation and （f, f）- interference law generation of knowledge, F-interference generation and F-interference law generation of two-directional S-rough sets are proposed. Based on the concepts above, the relation theoreras between F-interference loss and F-interference degree, the relation theorems between F-interference loss law and F-interference degree law, the dis- cernibility theorems between F-interference and F-interference law are presented. At last, the recognition criterion of F-interference law and its application are given.

Face recognition using illuminant locality preserving projections

A novel supervised manifold learning method was proposed to realize high accuracy face recognition under varying illuminant conditions. The proposed method, named illuminant locality preserving projections （ILPP）, exploited illuminant directions to alleviate the effect of illumination variations on face recognition. The face images were first projected into low dimensional subspace, Then the ILPP translated the face images along specific direction to reduce lighting variations in the face. The ILPP reduced the distance between face images of the same class, while increase the dis tance between face images of different classes. This proposed method was derived from the locality preserving projections （LPP） methods, and was designed to handle face images with various illumi nations. It preserved the face image＇ s local structure in low dimensional subspace. The ILPP meth od was compared with LPP and discriminant locality preserving projections （DLPP）, based on the YaleB face database. Experimental results showed the effectiveness of the proposed algorithm on the face recognition with various illuminations.

Quantitative recognition method for borehole trajectory models

In order to accurately calculate drilled trajectories,the method of quantitatively recognizing borehole trajectory models was provided,and a case analysis was conducted.Because the measurement-while-drilling data provide with measured values of tool-face angle besides inclination angle and azimuth angle,this paper presents the technological approach of recognizing borehole trajectory models based on tool-face angle.A universal tool-face angle equation was established based on the directional deflection mechanism of steerable drilling tools,and it can calculate the tool-face angles with characteristic parameters of various borehole trajectory models.Then,by evaluating the error between the theoretical values and the measured values of tool-face angle,the trajectory model most consistent with the actual well trajectory can be selected.The model recognition of borehole trajectory provides with the quantitative evaluation index and selection basis of survey calculation methods,which can avoid subjectively and randomly selecting the survey calculation method,and consequently improve the monitoring accuracy and reliability of borehole trajectory.

Multi Multi-Task Learning with Dynamic Splitting for Open Open-Set Wireless Signal Recognition

Open-set recognition(OSR)is a realistic problem in wireless signal recogni-tion,which means that during the inference phase there may appear unknown classes not seen in the training phase.The method of intra-class splitting(ICS)that splits samples of known classes to imitate unknown classes has achieved great performance.However,this approach relies too much on the predefined splitting ratio and may face huge performance degradation in new environment.In this paper,we train a multi-task learning(MTL)net-work based on the characteristics of wireless signals to improve the performance in new scenes.Besides,we provide a dynamic method to decide the splitting ratio per class to get more precise outer samples.To be specific,we make perturbations to the sample from the center of one class toward its adversarial direction and the change point of confidence scores during this process is used as the splitting threshold.We conduct several experi-ments on one wireless signal dataset collected at 2.4 GHz ISM band by LimeSDR and one open modulation recognition dataset,and the analytical results demonstrate the effective-ness of the proposed method.

融合人体骨架和姿势信息特征的轻量级人体动作识别方法

针对人体动作识别任务中特征值选取不当导致识别率低、使用多模态数据导致训练成本高等问题,提出一种轻量级人体动作识别方法。首先使用OpenPose、PoseNet提取出人体骨架信息,使用BWT69CL传感器提取姿势信息;其次对数据进行预处理、特征融合,对人体动作进行深度学习分类识别;最后,为验证此方法的有效性,在公开数据集WISDM、UCIHAR、HASC和自建的人体动作数据集上进行实验验证,并使用改进的目标引导注意力机制(target-guided attention,TGA)–长短期记忆(long short term memory,LSTM)网络输出最终的分类结果。实验结果表明,在自建数据集下融合姿势和骨架特征达到99.87%准确率,相比于只使用姿势信息特征,识别准确率提高了约5.31个百分点;相比于只使用人体骨架特征,识别准确率提高了约1.87个百分点;在识别时间上相比于只使用姿势信息,识别时间降低了约29.73 s;相比于只使用人体骨架数据,识别时间降低了约9 s。使用该方法能及时有效地反映人体的运动意图,有助于提高人体动作和行为的识别准确率和训练效率。

基于特征模式识别的农田监控系统研究

针对农田监控系统中存在的识别准确度不高、数据处理效率低等问题,提出了一种基于特征模式识别的农田监控系统。对农田监控系统的功能进行了设计,包括系统整体架构设计、数据采集与传输模块设计、特征提取与模式识别算法设计,结合用户界面设计实验与结果分析,验证了所提出系统的有效性和性能优势。对农田监控系统在实际农业生产中的应用前景进行了展望,并提出了未来改进方向和研究趋势。

基于标准和专利数据的企业技术研发方向识别研究

[研究目的]技术多样性逐渐标准化成为企业研发的关键目标,基于标准和专利的潜在关联识别企业研发方向可以降低企业研发不确定性,实现技术集成转化。[研究方法]提出一个以标准和专利数据为依据,集成标准文本及非文本信息解决研发方向识别问题的新方法。利用文本挖掘技术识别标准和专利信息,基于关键词向量链路预测模型寻找标准和专利的映射关系,构建关键词创新性评价模型,并对空白点进行综合评价,识别与评估研发方向,最后以远程医疗产业标准和专利数据为分析对象进行实证研究。[研究结论]实证结果表明:远程医疗产业识别出9个与专利技术具有较高相似性的标准空白点,通过进一步筛选过滤,最终确定的技术研发方向为:基于计算机网络平台的电子健康记录和在线咨询技术,基于5G和健康关怀技术的远程医疗护理技术,国际知名行业报告研究内容有力支撑了该识别结果。实证研究证明了所提出方法的可行性和可靠性,综合利用标准和专利之间的链接关系,集成标准文本及非文本信息可以进一步提升研发方向识别的效益,同时可以为产业整体创新发展和相关技术布局提供参考。

基于时空Inception残差注意力网络的脑电情绪识别

为了提高脑电情绪识别分类精度,最大限度利用脑电信号的空间和时间信息,提出一种Inception残差注意力卷积神经网络与双向长短期记忆(bi-directional long short-term memory, BiLSTM)网络相结合的新型架构时空Inception残差注意力网络。将脑电信号采集电极位置映射到二维矩阵中,采集信号作为通道,构成三维数据;将得到的三维数据输入到时空Inception残差注意力卷积网络之中,提取时空信息;将得到的特征输入到全连接层进行分类;将Inception结构引入脑电情绪识别领域,实现多尺度特征提取,并将电极映射到矩阵之中,保留电极位置信息,使用时空Inception残差注意力网络从时空两个维度获取脑电相关信息。实验表明,使用该模型对DEAP数据集进行情绪四分类可得到93.71%的准确度,相较于对比模型,识别精度提高了10%~20%。提出的模型在脑电信号情绪识别领域具有优良性能。

面向采购文件的跨模态图片文本命名实体识别

智慧供应链的数智化采购环节能够提高采购工作效率,节省大量人力成本。采购文件中包括大量证照资质等文件,针对其中图片文本中文字排版参差不齐、扫描图像不清晰等问题,设计了基于深度学习的端到端跨模态命名实体识别模型O2V2BLC(OCR-Vector-Bi-LSTM-CRF),从图片文本中识别命名实体。该模型针对光学字符识别技术识别出的图片文本字符,定义连续文本字符边界,将边界内每个文本字符映射为向量,设计双向长短期记忆(Bi-LSTM)网络捕获边界内字符序列的上下文语义,计算字符状态分数矩阵,并通过条件随机场约束字符标记序列规则,获得全局最优标记序列。针对训练集计算命名实体预测误差,动态优化O2V2BLC模型的参数,实现命名实体识别。将该方法应用于采购文件资质类型等图片文本数据,能够有效识别图片中的投标单位、专家姓名、专业名称等命名实体,与条件随机场、隐马尔可夫算法、Bert-Bi-LSTM-CRF模型进行对比,显著提高了实体识别准确率,为智慧供应链的数智化采购提供支持。

全频谱深度认知理论与方法

深入研究了全频谱认知与智能利用领域,强调了现有认知无线电理论和无线电监测技术在特定环境中的优势,指出在面对宽带、瞬变、未知、复杂的电磁环境时存在“认不深”的问题。为解决这一挑战,提出了多维度的深度频谱认知作为未来的发展方向,从信号多维特征与个体本征属性双重维度对辐射源目标进行精准认知。通过探索辐射源目标信号的多域特征,构建辐射源完备表征数据库,实现通信终端的本征属性认知,可以更全面地理解和利用频谱资源。分析了调制域、协议域中信号的调制识别以及协议解析等研究现状;高速目标信号识别和空域定向认知的发展瓶颈及主要研究方向。提供了较为全面的理论和方法,为未来多维度的电磁环境感知和频谱利用奠定了基础。

基于时序感知DAG的多模态对话情绪识别模型

针对现有对话情绪识别方法中对时序信息、话语者信息、多模态信息利用不充分的问题,提出了一个时序信息感知的多模态有向无环图模型(MTDAG)。其中所设计的时序感知单元能按照时间顺序优化话语权重设置,并收集历史情绪线索,实现基于近因效应下对时序信息和历史信息更有效的利用;设计的上下文和话语者信息融合模块,通过提取上下文语境和话语者自语境的深度联合信息实现对话语者信息的充分利用;通过设置DAG(directed acyclic graph)子图捕获多模态信息并约束交互方向的方式,在减少噪声引入的基础上充分利用多模态信息。在两个基准数据集IEMOCAP和MELD的大量实验表明该模型具有较好的情绪识别效果。

基于自注意力机制与词汇增强的中文医学命名实体识别

针对中文医学文本实体嵌套导致的单词边界识别困难问题以及现有栅格结构集成词汇特征所面临的语义信息损失严重的情况,提出一种用于中文医学命名实体识别(MNER)的自适应词汇信息增强模型。首先,利用双向长短期记忆(BiLSTM)网络编码字符序列的上下文信息并捕捉较长距离的依赖关系;然后,对字符序列中每个字符的潜在单词信息进行字词对建模,采用自注意力机制实现不同单词之间的内部交互;最后,通过基于双线性注意力机制的词汇适配器将词汇信息集成到文本序列中的每个字符中,有效增强语义信息的同时充分利用单词丰富的边界信息,并抑制相关性低的单词。实验结果表明,所提模型与基于字符的基线模型相比,平均F1值分别提升了1.37~2.38个百分点,并在结合BERT后取得了最优的效果。

基于direct LDA的幅度谱子空间雷达目标识别

针对高分辨距离像(HRRP)可分性低和维数高的问题,提出一种新的雷达自动目标识别(RATR)方法:dLDA&SVM。先采用直接线性判别分析在HRRP的幅度谱空间进行特征提取,然后在子空间中采用角域均值模板库训练one-against-all支撑向量机(SVM)多类分类器进行目标识别。并设计了最短距离分类器与SVM分类器比较。基于外场实测数据的实验结果表明,与LDA幅度谱子空间法,幅度谱原空间法相比,dLDA&SVM可显著降低数据维数并提高识别性能。

多尺度特征与知识图谱融合的景区手写诗词识别

针对景区手写诗词存在背景纹理复杂、字体尺寸及风格多样等特点导致景区游客难以识别手写诗词的问题,首先,分析研究景区手写诗词的识别场景,设计景区诗词检测网络(detection of poetry in scenic areas-network,DPSA-Net)以提取景区手写诗词不同尺度的特征,并结合手写诗词字符间的链接依赖关系实现景区手写诗词检测;其次,设计了卷积循环聚合网络(convolution recurrent aggregation network,CRA-Net)以对景区手写诗词进行识别,结合卷积神经网络(convolutional neural networks,CNN)和双向长短期记忆网络提取手写诗词图像的序列特征,并通过聚合交叉熵(aggregation cross-entropy,ACE)实现特征向文本的转换;最后,结合景区知识图谱对CRA-Net的输出进行校正,进而提高景区手写诗词的识别准确率。实验结果表明,通过景区手写诗词矫正技术对CRA-Net的识别结果矫正后,识别准确率达到了79.04%,同时,该技术具有较好的抗干扰能力和良好的应用前景。