基于注意力机制与多尺度特征融合的行人重识别方法

针对行人重识别中因遮挡、姿态变化使模型特征无法充分表达行人信息的问题,提出了基于注意力机制与多尺度特征融合的行人重识别方法.首先使用改进的骨干网络R-ResNet50提取图像特征;其次,抽取网络不同尺度的特征层嵌入注意力机制DANet,使模型更关注于重点信息;最后,对提取出的关键特征进行多尺度特征融合,实现特征间的优势...     

面向智能船舶的水面小目标检测算法

针对智能船舶中基于视觉传感器的水面小目标识别具有识别区域分辨率低、图像模糊、信噪比低等问题,提出了一种新的基于卷积神经网络的水面小目标检测算法——自注意力特征融合检测算法.首先,为了提高视觉信息处理的效率与准确性,在网络模型中引入了自注意力模块,更多关注小目标的细节信息.其次,在网络模型中采用了结构化的特征融合算法,通过多尺度语义信息融合提升对小目标的检测性能.最后,为了解决目标检测的定位问题,在smooth L1损失函数的基础上设计了一种大梯度定位损失函数.通过与传统的Faster R-CNN目标检测算法在船舶数据库上进行仿真对比,验证了所提算法在解决水面小目标检测问题上的有效性.     

基于DenseNet的人脸图像情绪识别研究

针对人脸情绪识别类内差异大,类间差异小的特点,结合学生人脸图像的线上课堂情绪识别的场景,提出多尺度空洞卷积模块提取不同空间尺度特征的稠密深度神经网络模型,实现自然场景下学生人脸图像识别.该模型主要由多尺度空洞卷积和DenseNet神经网络两个子网络组成,其中多尺度空洞卷积由不同空洞率的四分支网络提取不同尺度特征,空洞卷积减小特征图尺寸,减少DenseNet内存资源占用;最后在DenseNet网络中结合Adam优化器和中心损失函数.使用稠密网络的旁路连接,加强情绪特征传递和复用.研究结果表明：基于稠密深度神经网络的情绪识别网络模型能够有效提高情绪分类的准确率,模型对预处理后的FER2013+数据集识别准确率达到93.99%,可为线上教学反馈提供技术支持.     

基于多尺度特征融合的单目图像深度估计

为解决从单目图像中很难恢复出准确、有效深度信息的问题,提出一种多尺度特征融合的单目图像深度估计算法.算法采用端对端训练的卷积神经网络(CNN)结构,引入从图像编码器到解码器的跳层连接来实现在不同尺度上特征的提取和表达,设计了一种多尺度的损失函数来提升卷积神经网络的训练效果.通过在NYU Depth V2室内场景深度数据集和KITTI室外场景深度数据集上的训练、验证和测试,实验结果表明:提出的多尺度特征融合方法得到的深度图边缘清晰、层次分明,且在室内场景和室外场景中均能适用,具有较强的泛化性,可以适应多种实际场景的需求.     

基于图卷积的跨模态行人重识别方法

近年来,红外-可见光的行人重识别在视频监控、网络刑侦等领域应用广泛,这项任务的目的是实现RGB摄像机和红外摄像机下出现的同一行人的匹配。由于行人图像在RGB模态和红外模态下存在较大差异,因而使得该项任务具有一定的挑战性。文中提出一种基于图卷积的跨模态行人重识别方法,同时提出一种新颖的异心三元组损失函数,用于更好表征行人特征。该方法首先对水平切割方法进行改进,在此基数上以局部特征和全局特征为节点构建图卷积神经网络,并利用构建的图卷积神经网络学习图像结构化特征;然后,引入了一种全新的异心三元组损失函数,并结合Softmax损失函数进一步提高模型性能。两个公开数据集上进行的对比实验、消融实验以及可视化实验结果验证了文中所提方法的卓越性能。     

面向行人重识别分集特征挖掘的多分支深度神经网络构造进展

随着深度学习的兴起,近年来行人重识别性能提高的关键是如何构建具有分集特征挖掘的深度神经网络以获得行人图像的紧凑表达,而多分支深度神经网络是实现分集特征挖掘的主要途径.文中分析了近年来用于行人重识别多分支深度神经网络的几种主流构建方法,包括局部分割分集、注意力分集、特征遮挡分集、异构支路分集等,并在3个行人重识别数据集M...     

基于注意力特征融合的SqueezeNet细粒度图像分类模型

针对现有细粒度图像分类算法普遍存在的模型结构复杂、参数多、分类准确率较低等问题,提出一种注意力特征融合的SqueezeNet细粒度图像分类模型.通过对现有细粒度图像分类算法和轻量级卷积神经网络的分析,首先使用3个典型的预训练轻量级卷积神经网络,对其微调后在公开的细粒度图像数据集上进行验证,经比较后选择了模型性能最佳的SqueezeNet作为图像的特征提取器;然后将两个具有注意力机制的卷积模块嵌入至SqueezeNet网络的每个Fire模块;接着提取出改进后的SqueezeNet的中间层特征进行双线性融合形成新的注意力特征图,与网络的全局特征再融合后分类;最后通过实验对比和可视化分析,网络嵌入Convolution Block Attention Module(CBAM)模块的分类准确率在鸟类、汽车、飞机数据集上依次提高了8.96%、4.89%和5.85%,嵌入Squeeze-and-Excitation(SE)模块的分类准确率依次提高了9.81%、4.52%和2.30%,且新模型在参数量、运行效率等方面比现有算法更具优势.     

基于YOLOv3的改进行人检测算法研究

针对将YOLOv3通用目标检测算法应用于行人检测时的检测精度低、定位不准确的问题,提出了一种基于YOLOv3的适用于行人体态特征的目标检测算法。在预处理生成先验框部分,将MSCOCO通用数据集改进为MSCOCO中的person子集来生成仅针对行人体态特征的锚框,并将生成先验框的K-means算法改进为K-means++算法以弥补K-means算法选择中心点的随意性。针对特征提取,改进了卷积神经网络(CNN) Darknet-53的结构以使其能够提取到更多与行人尺度有关的特征信息。损失函数中的目标定位部分改进为更加符合行人体态特征的损失函数。实验使用MSCOCO训练集中的person子集训练模型,分别使用MSCOCO测试集的person子集和自制的行人数据集作为测试集验证模型。结果显示,对比YOLOv3的通用目标检测算法,该文改进算法更易于提取行人特征并提供有效反馈,提高了检测精度。     

基于多层次深度学习网络的行人重识别

提出了多层级特征融合模型,该模型利用深度学习网络提取行人图像的全局特征和局部特征,并将全局和局部特征联合起来,以生成更具辨识度的描述符.在模型中,基于部分的多层级网络用于提取不同网络深度的局部特征,从而将网络底层到高层中提取的局部特征组合起来.全局—局部网络分支则提取网络深层的局部特征和全局特征,用于识别行人.该模型在三个数据集上进行了实验并得到了更好的结果.     

结合注意力机制与时空特征融合的长时程行为识别方法

为了建立基于视频行为识别的长时程图像序列的时空信息模型,文章提出了一种嵌入注意力的时空特征融合网络(attention-embedded spatial-temporal feature fusion network, ASTFFN)的深度神经网络模型。ASTFFN将一个包含动作的长时程图像序列分割成多个重叠的片段,并使用一个注意力嵌入特征提取网络(attention-embedded feature extraction network, AFEN)从每个片段中的RGB图像或光流图像中提取经过注意力加权的时空特征,进而融合每个片段的时空加权特征,生成行为识别的视频级预测。基于UCF101和HMDB51行为识别数据库进行了大量的实验,验证了方法的有效性。与目前主流的行为识别算法相比,该文方法在识别精度上取得了较好的效果。     

Markedness and its Interpretation in Language Use

Language markedness is a common phenomenon in languages, and is reflected from hearing, vision and sense, i.e. the variation in the three aspects such as phonology, morphology and semantics. This paper focuses on the interpretation of markedness in language use following the three perspectives, i.e. pragmatic interpretation, psychological interpretation and cognitive interpretation, with an aim to define the function of markedness.     

Lower Paleozoic oil relationships within Williston Basin, Canada

The Williston Basin is a significant petroleum province, containing oil production zones that include the Middle Cambrian to Lower Ordovician, Upper Ordovician, Middle Devonian, Upper Devonian and Mississippian and within the Jurassic and Cretaceous. The oils of the Williston Basin exhibit a wide range of geochemical characteristics defined as "oil families", although the geochemical signature of the Cambrian Deadwood Formation and Lower Ordovician Winnipeg reservoired oils does not match any "oil family". Despite their close stratigraphic proximity, it is evident that the oils of the Lower Palaeozoic within the Williston Basin are distinct. This suggests the presence of a new "oil family" within the Williston Basin. Diagnostic geochemical signatures occur in the gasoline range chromatograms, within saturate fraction gas chromatograms and biomarker fingerprints. However, some of the established criteria and cross-plots that are currently used to segregate oils into distinct genetic families within the basin do not always meet with success, particularly when applied to the Lower Palaeozoic oils of the Deadwood and Winnipeg Formation.     

An Analysis of Catherine＇s Character——Wuthering Heights

Wuthering Heights, Emily Bronte＇s only novel, was published in December of 1847 under the pseudonym Ellis Bell. The book did not gain immediate success, but it is now thought one of the finest novels in the English language. Catherine is the key character of this masterpiece, because everybody and everything center on her though she had a short life. We can understand this masterpiece better if we know Catherine well.     

Characterization of Ordovician carbonate reservoirs, southeastern Saskatchewan, Canada

The discovery of the prolific Ordovician Red River reservoirs in 1995 in southeastern Saskatchewan was the catalyst for extensive exploration activity which resulted in the discovery of more than 15 new Red River pools. The best yields of Red River production to date have been from dolomite reservoirs. Understanding the processes of dolomitization is, therefore, crucial for the prediction of the connectivity, spatial distribution and heterogeneity of dolomite reservoirs.The Red River reservoirs in the Midale area consist of 3～4 thin dolomitized zones, with a total thickness of about 20 m, which occur at the top of the Yeoman Formation. Two types of replacement dolomite were recognized in the Red River reservoir: dolomitized burrow infills and dolomitized host matrix. The spatial distribution of dolomite suggests that burrowing organisms played an important role in facilitating the fluid flow in the backfilled sediments. This resulted in penecontemporaneous dolomitization of burrow infills by normal seawater. The dolomite in the host matrix is interpreted as having occurred at shallow burial by evaporitic seawater during precipitation of Lake Almar anhydrite that immediately overlies the Yeoman Formation. However, the low δ18O values of dolomited burrow infills (-5.9‰～ -7.8‰, PDB) and matrix dolomites (-6.6‰～ -8.1‰, avg. -7.4‰ PDB) compared to the estimated values for the late Ordovician marine dolomite could be attributed to modification and alteration of dolomite at higher temperatures during deeper burial, which could also be responsible for its 87Sr/86Sr ratios (0.7084～0.7088) that are higher than suggested for the late Ordovician seawaters (0.7078～0.7080). The trace amounts of saddle dolomite cement in the Red River carbonates are probably related to "cannibalization" of earlier replacement dolomite during the chemical compaction.     

Features of Women's language

Language is a means of verbal communication. People use language to communicate with each other. In the society, no two speakers are exactly alike in the way of speaking. Some differences are due to age, gender, statue and personality. Above all, gender is one of the obvious reasons. The writer of this paper tries to describe the features of women＇s language from these perspectives： pronunciation, intonation, diction, subjects, grammar and discourse. From the discussion of the features of women＇s language, more attention should be paid to language use in social context. What＇s more, the linguistic phenomena in a speaking community can be understood more thoroughly.     

Design and analysis of generic LBS application developing platform

Location based services is promising due to its novel working style and contents.A software platform is proposed to provide application programs of typical location based services and support new applications developing efficiently. The analysis shows that this scheme is easy implemented, low cost and adapt to all kinds of mobile nework system.     

橡胶颗粒沥青混合料低温抗裂性试验研究

以AC-13级配为基础,将橡胶颗粒代替部分集料掺入混合料中,以低温弯曲试验为评价方法对不同橡胶颗粒掺量下沥青混合料的低温抗裂性进行研究,并引入应变能密度值对混合料的低温抗裂性进行综合评价.试验结果表明:橡胶颗粒沥青混合料试件的破坏微应变均超过2 300,满足冬寒区的技术指标;无论是否掺加橡胶颗粒,随着温度的下降,沥青混合料破坏时的最大弯拉强度增大,弯拉应变降低,劲度模量增大;弯曲应变能密度在胶粒掺量为1%左右时具有较大的弯曲应变能密度值,此时橡胶颗粒沥青混合料具有较好的低温抗裂性.     

A computer generator for randomly layered structures

AcomputergeneratorforrandomlylayeredstructuresYUJia shun1,2,HEZhen hua2(1.TheInstituteofGeologicalandNuclearSciences,NewZealand;2.StateKeyLaboratoryofOilandGasReservoirGeologyandExploitation,ChengduUniversityofTechnology,China)Abstract:Analgorithmisintrod…     

低渗透非均质砂岩油藏启动压力梯度研究

理论推导与室内实验相结合,建立了低渗透非均质砂岩油藏启动压力梯度确定方法。首先借助油藏流场与电场相似的原理,推导了非均质砂岩油藏启动压力梯度计算公式。其次基于稳定流实验方法,建立了非均质砂岩油藏启动压力梯度测试方法。结果表明:低渗透非均质砂岩油藏的启动压力梯度确定遵循两个等效原则。平面非均质油藏的启动压力梯度等于各级渗透率段的启动压力梯度关于长度的加权平均;纵向非均质油藏的启动压力梯度等于各渗透率层的启动压力梯度关于渗透率与渗流面积乘积的加权平均。研究成果可用于有效指导低渗透非均质砂岩油藏的合理井距确定,促进该类油藏的高效开发。     

Quantitative trait loci （QTLs） for quality traits related to protein and starch in wheat

Quality traits in wheat （Triticum aestirum L.） were studied by quantitative trait locus （QTL） analysis in a recombinant inbred line （RIL） population, a set of 131 lines derived from Chuan 35050 × Shannong 483 cross （ChSh）. Grains from RILs were assayed for 21 quality traits related to protein and starch. A total of 35 putative QTLs for 19 traits with a single QTL explaining 7.99-40.52% of phenotypic variations were detected on 10 chromosomes, 1D, 2A, 2D, 3B, 3D, 5A, 6A, 6B, 6D, and 7B. The additive effects of 30 QTLs were positive, contributed by Chuan 35050, the remaining 5 QTLs were negative with the additive effect contributed by Shannong 483. For protein traits, 15 QTLs were obtained and most of them were located on chromosomes 1 D, 3B and 6D, while 20 QTLs for starch traits were detected and most of them were located on chromosomes 3D, 6B and 7B. Only 7 QTLs for protein and starch traits were co-located in three regions on chromosomes 1D, 2A and 2D. These protein and starch trait QTLs showed a distinct distribution pattern in certain regions and chromosomes. Twenty-two QTLs were clustered in 6 regions of 5 chromosomes. Two QTL clusters for protein traits were located on chromosomes 1D and 3B, respectively, three clusters for starch traits on chromosomes 3D, 6B and 7B, and one cluster including protein and starch traits on chromosome 1D.