Chengdu teahouses,as core public spaces in marketplace society,have undergone transformative reconstruction-from“containers of everyday life”to“containers of commercial traffic and digital flows”-during the proces...Chengdu teahouses,as core public spaces in marketplace society,have undergone transformative reconstruction-from“containers of everyday life”to“containers of commercial traffic and digital flows”-during the process of modernization.Employing spatial archaeology as a methodology,combined with fieldwork and analysis of historical documents,this study systematically examines the diachronic evolution of architectural forms,functional orientations,and social networks within Chengdu teahouses.The study reveals the logic of spatial reconstruction under the interplay of multiple forces,including cultural heritage preservation,capital-driven development,and technological intervention.The findings identify three paradigms of spatial transformation in teahouses.First,heritage specimenization,which reinforces the continuity of collective memory through symbolic extraction but risks diminishing the vitality of everyday social interactions.Second,consumption upgrading,which caters to the demands of emerging groups through iterative business models yet necessitates vigilance against spatial differentiation eroding marketplace inclusivity.Third,digital parasitism,which expands communicative dimensions through technological empowerment but confronts the risk of flattening localized knowledge.These paradigms reflect both adaptive responses of traditional spaces to contemporary pressure and the tension of reconstruction imposed by instrumental rationality on marketplace networks.The study demonstrates that spatial transformation in Chengdu teahouses is not unidirectional alienation but rather a multifaceted configuration where the continuity of tradition coexists with innovative practices amid functional diversification.This research advocates for striking a balance between the preservation of traditional spaces and modern renewal and explores organic integration approaches for traditional and modern elements,thereby providing a theoretical framework and practical insights for the transformation of traditional public spaces.展开更多
The rapid development of digital technologies has driven the emergence and popularization of online-to-offline(O2O)retail,reshaping the retail landscape in urban China.However,spatial distribution characteristics and ...The rapid development of digital technologies has driven the emergence and popularization of online-to-offline(O2O)retail,reshaping the retail landscape in urban China.However,spatial distribution characteristics and influencing mechanisms of emerging O2O retail have not been thoroughly investigated in extant studies.Taking the central urban area of Guangzhou as the case,this study utilized multi-source data and machine learning methods to explore the distribution characteristics of O2O retail space and to further identify the nonlinear effects of the built environment,sociodemographic,and economic factors on its distribution.The results revealed that O2O retail space exhibited a‘single-center’distribution pattern,in contrast to the‘multi-center’distribution pattern of traditional retail space.This finding supported the diffusion of innovation hypothesis,highlighting that the expansion of O2O retail modes first spread from traditional developed retail space.Furthermore,spatial heterogeneities were observed across different types of O2O retail space,with O2O in-store showing a‘core-periphery’spatial structure as described by Central Place Theory,whereas O2O delivery displaying a‘horizontal,non-hierarchical,and multi-centered’network structure following Central Flow Theory.Compared to traditional retail space,the distribution of O2O retail space was more influenced by sociodemographic factors such as the proportion of youth,education level,and income level,but less affected by the built environment factors like office and building density.Furthermore,nonlinear effects of these influencing factors on the distribution of O2O retail space were identified,which enriched the existing literature by highlighting effective ranges and threshold effects.These findings provided valuable insights into O2O retail space development in the context of digital transformation.展开更多
Microphone array-based sound source localization(SSL)is a challenging task in adverse acoustic scenarios.To address this,a novel SSL algorithm based on deep neural network(DNN)using steered response power-phase transf...Microphone array-based sound source localization(SSL)is a challenging task in adverse acoustic scenarios.To address this,a novel SSL algorithm based on deep neural network(DNN)using steered response power-phase transform(SRP-PHAT)spatial spectrum as input feature is presented in this paper.Since the SRP-PHAT spatial power spectrum contains spatial location information,it is adopted as the input feature for sound source localization.DNN is exploited to extract the efficient location information from SRP-PHAT spatial power spectrum due to its advantage on extracting high-level features.SRP-PHAT at each steering position within a frame is arranged into a vector,which is treated as DNN input.A DNN model which can map the SRP-PHAT spatial spectrum to the azimuth of sound source is learned from the training signals.The azimuth of sound source is estimated through trained DNN model from the testing signals.Experiment results demonstrate that the proposed algorithm significantly improves localization performance whether the training and testing condition setup are the same or not,and is more robust to noise and reverberation.展开更多
The Yangtze River Delta(YRD) is a region in China with a serious contradiction between economic growth and environmental pollution. Exploring the spatiotemporal effects and influencing factors of air pollution in the ...The Yangtze River Delta(YRD) is a region in China with a serious contradiction between economic growth and environmental pollution. Exploring the spatiotemporal effects and influencing factors of air pollution in the region is highly important for formulating policies to promote the high-quality development of urban industries. This study uses the spatial Durbin model(SDM) to analyze the local direct and spatial spillover effects of industrial transformation on air pollution and quantifies the contribution of each factor. From 2008 to 2018, there was a significant spatial agglomeration of industrial sulfur dioxide emissions(ISDE) in the YRD, and every 1% increase in ISDE led to a synchronous increase of 0.603% in the ISDE in adjacent cities. The industrial scale index(ISCI) and industrial structure index(ISTI), as the core factors of industrial transformation, significantly affect the emissions of sulfur dioxide in the YRD, and the elastic coefficients are 0.677 and-0.368, respectively. The order of the direct effect of the explanatory variables on local ISDE is ISCI>ISTI>foreign direct investment(FDI)>enterprise technological innovation(ETI)>environmental regulation(ER)> per capita GDP(PGDP). Similarly, the order of the spatial spillover effect of all variables on ISDE in adjacent cities is ISCI>PGDP>FDI>ETI>ISTI>ER, and the coefficients of the ISCI and ISTI are 1.531 and 0.113, respectively. This study contributes to the existing research that verifies the environmental Kuznets curve in the YRD, denies the pollution heaven hypothesis, indicates the Porter hypothesis, and provides empirical evidence for the formation mechanism of regional environmental pollution from a spatial spillover perspective.展开更多
Blade-health monitoring is intensely required for turbomachinery because of the high failure risk of rotating blades.Blade-Tip Timing(BTT)is considered as the most promising technique for operational blade-vibration m...Blade-health monitoring is intensely required for turbomachinery because of the high failure risk of rotating blades.Blade-Tip Timing(BTT)is considered as the most promising technique for operational blade-vibration monitoring,which obtains the parameters that characterize the blade condition from recorded signals.However,its application is hindered by severe undersampling and stringent probe layouts.An inappropriate probe layout can make most of the existing methods invalid or inaccurate.Additionally,a general conflict arises between the allowed and required layouts because of arrangement restrictions.For the sake of economy and safety,parameter identification based on fewer probes has been preferred by users.In this work,a spatial-transformation-based method for parameter identification is proposed based on a single-probe BTT measurement.To present the general Sampling-Aliasing Frequency(SAFE)map definition,the traditional time-frequency analysis methods are extended to a time-sampling frequency.Then,a SAFE map is projected onto a parameter space using spatial transformation to extract the slope and intercept parameters,which can be physically interpreted as an engine order and a natural frequency using coordinate transformation.Finally,the effectiveness and robustness of the proposed method are verified by simulations and experiments under uniformly and nonuniformly variable speed conditions.展开更多
【目的】高分辨率遥感影像语义分割通过精准提取地物信息,为城市规划、土地分析利用提供了重要的数据支持。当前分割方法通常将遥感影像划分为标准块,进行多尺度局部分割和层次推理,未充分考虑影像中的上下文先验知识和局部特征交互能力...【目的】高分辨率遥感影像语义分割通过精准提取地物信息,为城市规划、土地分析利用提供了重要的数据支持。当前分割方法通常将遥感影像划分为标准块,进行多尺度局部分割和层次推理,未充分考虑影像中的上下文先验知识和局部特征交互能力,影响了推理分割质量。【方法】为了解决这一问题,本文提出了一种联合跨尺度注意力和语义视觉Transformer的遥感影像分割框架(Cross-scale Attention Transformer,CATrans),融合跨尺度注意力模块和语义视觉Transformer,提取上下文先验知识增强局部特征表示和分割性能。首先,跨尺度注意力模块通过空间和通道两个维度进行并行特征处理,分析浅层-深层和局部-全局特征之间的依赖关系,提升对遥感影像中不同粒度对象的注意力。其次,语义视觉Transformer通过空间注意力机制捕捉上下文语义信息,建模语义信息之间的依赖关系。【结果】本文在DeepGlobe、Inria Aerial和LoveDA数据集上进行对比实验,结果表明:CATrans的分割性能优于现有的WSDNet(Discrete Wavelet Smooth Network)和ISDNet(Integrating Shallow and Deep Network)等分割算法,分别取得了76.2%、79.2%、54.2%的平均交并比(Mean Intersection over Union,mIoU)和86.5%、87.8%、66.8%的平均F1得分(Mean F1 Score,mF1),推理速度分别达到38.1 FPS、13.2 FPS和95.22 FPS。相较于本文所对比的最佳方法WSDNet,mIoU和mF1在3个数据集中分别提升2.1%、4.0%、5.3%和1.3%、1.8%、5.6%,在每类地物的分割中都具有显著优势。【结论】本方法实现了高效率、高精度的高分辨率遥感影像语义分割。展开更多
Relative to hospitalized patient information, outpatient admission information is relatively simple. It only includes the patient admission time, place of residence and other information. Traditionally, the excavation...Relative to hospitalized patient information, outpatient admission information is relatively simple. It only includes the patient admission time, place of residence and other information. Traditionally, the excavation of this information is not sufficient. However, when a large number of patients admitted time and residence information combined to consider, and add some data mining technology, some of the previously ignored regular information is likely to be found. Using 5 years of data mining research and admission data from a paediatric department at a large women’s and children’s hospital in China, we found important fluctuation rules regarding admissions using wavelet analysis on hospital admission data among different scales of cyclical fluctuations. Method: Seasonal distribution of patient number was analysed based on Haar wavelet transformation, and level 3 and level 2 of wavelets were extracted out to fit the data. The distribution function of hospitalized patients was visualized by kernel density estimation. Using linear regression and ARIMA (autoregressive integrated moving average model) predict the seasonally number of patients in the future. Results: The data analysis demonstrates the total surge of inpatients was decomposed into one mother wavelet and five small wavelets, each of which represents different time frequency. Besides, as distance from hospital increases, the number of patients decreased exponentially. The seasonal factors are the largest time factor influencing the number changes of patients. Conclusion: By wavelet analysis and the improved prediction model, we could make forecast on the future inpatient number trend and prove factors such as geographic position is influential on inpatient amount. Additionally, the concept of data mining based on spatial distribution and spectral analysis could be applied to other aspects of social management.展开更多
By introducing an optic-null medium into the finite embedded transformation,a reflectionless spatial beam bender is designed,which can steer the output beam by a fixed pre-designed angleβfor an arbitrary incident ang...By introducing an optic-null medium into the finite embedded transformation,a reflectionless spatial beam bender is designed,which can steer the output beam by a fixed pre-designed angleβfor an arbitrary incident angle.The bending angleβof the beam bender is determined by the geometrical angle of the device,which can be changed by simply choosing different geometrical angles.For various bending angles,the designed spatial beam bender can be realized by the same materials(i.e.,an optic-null medium),which is a homogenous anisotropic material.Numerical simulations verify the reflectionless bending effect and rotated imaging ability of the proposed beam bender.A reduction model of the optic-null medium is studied,which can also be used for a reflectionless spatial beam bender with a pre-designed bending angle.展开更多
Eddy current pulsed thermography（ECPT） is an emerging Non-destructive testing and evaluation（NDT E） technique, which uses hybrid eddy current and thermography NDT E techniques that enhances the detectability fro...Eddy current pulsed thermography（ECPT） is an emerging Non-destructive testing and evaluation（NDT E） technique, which uses hybrid eddy current and thermography NDT E techniques that enhances the detectability from their compensation. Currently, this technique is limited by the manual selection of proper contrast frames and the issue of improving the efficiency of defect detection of complex structure samples remains a challenge. In order to select a specific frame from transient thermal image sequences to maximize the contrast of thermal variation and defect pattern from complex structure samples, an energy driven approach to compute the coefficient energy of wavelet transform is proposed which has the potential of automatically selecting both optimal transient frame and spatial scale for defect detection using ECPT. According to analysis of the variation of different frequency component and the comparison study of the detection performance of different scale and wavelets, the frame at the end of heating phase is automatically selected as an optimal transient frame for defect detection. In addition, the detection capabilities of the complex structure samples can be enhanced through proper spatial scale and wavelet selection. The proposed method has successfully been applied to low speed impact damage detection of carbon fibre reinforced polymer（CFRP） composite as well as providing the guidance to improve the detectability of ECPT technique.展开更多
This letter exploits fundamental characteristics of a wavelet transform image to form a progressive octave-based spatial resolution. Each wavelet subband is coded based on zeroblock and quardtree partitioning ordering...This letter exploits fundamental characteristics of a wavelet transform image to form a progressive octave-based spatial resolution. Each wavelet subband is coded based on zeroblock and quardtree partitioning ordering scheme with memory optimization technique. The method proposed in this letter is of low complexity and efficient for Internet plug-in software.展开更多
The spatial transformations can be observed at different religious-historic towns of India due to urbanization. Research is based upon fact that there is substantial change in the built environment because of spatial ...The spatial transformations can be observed at different religious-historic towns of India due to urbanization. Research is based upon fact that there is substantial change in the built environment because of spatial transformations at the religious-historic towns. The process of modernization in the functions and spatial layout is unavoidable at any historic town. The study attempts to focus on various urban historic conservation components, including the look of historic buildings, their earlier uses, and its immediate surroundings to improve the built environment of historic towns. A theoretical framework for the urban conservation of ancient towns is the main objective of study. How to modernize the historic conservation function while preserving the space’s texture and integrity. The research started with the investigation of the morphological growth of Mathura district, India through satellite images and in-depth study of the evolution process of street network in Vrindavan town, which is one of the main temple towns of Mathura district. There is a significant difference in the layout & architectural character of old part and the newly developed Vrindavan. Due to increased accessibility and movement, the spatial structure of traditional religious precincts, which were once local integration centres, has significantly changed. Increasing & changing mode of transportation and further increase in the religious tourism might be the cause or a big reason for the spatial transformations and correspondingly there is a challenge to conserve & preserve the religious precincts of historic towns. The study tries to analyze spatial transformations with the help of Historical GIS at different scales of urban form. Suggestive measures to conserve the environmental ambience of religious-historic towns are the outcome of the research.展开更多
本文提出了一种基于双交叉注意力融合的Swin-AK Transformer(Swin Transformer based on alterable kernel convolution)和手工特征相结合的智能手机拍摄图像质量评价方法。首先,提取了影响图像质量的手工特征,这些特征可以捕捉到图像...本文提出了一种基于双交叉注意力融合的Swin-AK Transformer(Swin Transformer based on alterable kernel convolution)和手工特征相结合的智能手机拍摄图像质量评价方法。首先,提取了影响图像质量的手工特征,这些特征可以捕捉到图像中细微的视觉变化;其次,提出了Swin-AK Transformer,增强了模型对局部信息的提取和处理能力。此外,本文设计了双交叉注意力融合模块,结合空间注意力和通道注意力机制,融合了手工特征与深度特征,实现了更加精确的图像质量预测。实验结果表明,在SPAQ和LIVE-C数据集上,皮尔森线性相关系数分别达到0.932和0.885,斯皮尔曼等级排序相关系数分别达到0.929和0.858。上述结果证明了本文提出的方法能够有效地预测智能手机拍摄图像的质量。展开更多
Background Several face detection and recogni tion methods have been proposed in the past decades that have excellent performance.The conventional face recognition pipeline comprises the following:(1)face detection,(2...Background Several face detection and recogni tion methods have been proposed in the past decades that have excellent performance.The conventional face recognition pipeline comprises the following:(1)face detection,(2)face alignment,(3)feature extraction,and(4)similarity,which are independent of each other.The separate facial analysis stages lead to redundant model calculations,and are difficult for use in end-to-end training.Methods In this paper,we propose a novel end-to-end trainable convolutional network framework for face detection and recognition,in which a geometric transformation matrix is directly learned to align the faces rather than predicting the facial landmarks.In the training stage,our single CNN model is supervised only by face bounding boxes and personal identities,which are publicly available from WIDER FACE and CASIA-WebFace datasets.Our model is tested on Face Detection Dataset and Benchmark(FDDB)and Labeled Face in the Wild(LFW)datasets.Results The results show 89.24%recall for face detection tasks and 98.63%accura cy for face recognition tasks.展开更多
提出了一种Transformer与图网络相结合的网络模型,用于对视觉传感器采集到的视频图像进行三维人体姿态估计。Transformer能够有效地从二维关键关节点中提取时空维度高相关性特征,而图网络则能够感知细节相关性特征,通过融合这两种网络结...提出了一种Transformer与图网络相结合的网络模型,用于对视觉传感器采集到的视频图像进行三维人体姿态估计。Transformer能够有效地从二维关键关节点中提取时空维度高相关性特征,而图网络则能够感知细节相关性特征,通过融合这两种网络结构,提高了三维姿态估计的精度。在公开数据集Human3.6M上进行了仿真实验,验证了Transformer与图卷积融合算法的性能。实验结果显示,最终估计得到的三维人体关节点的平均关节点位置偏差(Mean Per Joint Position Error,MPJPE)为38.4 mm,相较于现有方法有一定提升,表明该方法具有较强的应用价值,可应用于许多下游相关工作中。展开更多
In view of the weak ability of the convolutional neural networks to explicitly learn spatial invariance and the probabilistic loss of discriminative features caused by occlusion and background interference in pedestri...In view of the weak ability of the convolutional neural networks to explicitly learn spatial invariance and the probabilistic loss of discriminative features caused by occlusion and background interference in pedestrian re-identification tasks,a person re-identification method combining spatial feature learning and multi-granularity feature fusion was proposed.First,an attention spatial transformation network(A-STN)is proposed to learn spatial features and solve the problem of misalignment of pedestrian spatial features.Then the network was divided into a global branch,a local coarse-grained fusion branch,and a local fine-grained fusion branch to extract pedestrian global features,coarse-grained fusion features,and fine-grained fusion features,respectively.Among them,the global branch enriches the global features by fusing different pooling features.The local coarse-grained fusion branch uses an overlay pooling to enhance each local feature while learning the correlation relationship between multi-granularity features.The local fine-grained fusion branch uses a differential pooling to obtain the differential features that were fused with global features to learn the relationship between pedestrian local features and pedestrian global features.Finally,the proposed method was compared on three public datasets:Market1501,DukeMTMC-ReID and CUHK03.The experimental results were better than those of the comparative methods,which verifies the effectiveness of the proposed method.展开更多
To enable proper diagnosis of a patient,medical images must demonstrate no presence of noise and artifacts.The major hurdle lies in acquiring these images in such a manner that extraneous variables,causing distortions...To enable proper diagnosis of a patient,medical images must demonstrate no presence of noise and artifacts.The major hurdle lies in acquiring these images in such a manner that extraneous variables,causing distortions in the form of noise and artifacts,are kept to a bare minimum.The unexpected change realized during the acquisition process specifically attacks the integrity of the image’s quality,while indirectly attacking the effectiveness of the diagnostic process.It is thus crucial that this is attended to with maximum efficiency at the level of pertinent expertise.The solution to these challenges presents a complex dilemma at the acquisition stage,where image processing techniques must be adopted.The necessity of this mandatory image pre-processing step underpins the implementation of traditional state-of-the-art methods to create functional and robust denoising or recovery devices.This article hereby provides an extensive systematic review of the above techniques,with the purpose of presenting a systematic evaluation of their effect on medical images under three different distributions of noise,i.e.,Gaussian,Poisson,and Rician.A thorough analysis of these methods is conducted using eight evaluation parameters to highlight the unique features of each method.The covered denoising methods are essential in actual clinical scenarios where the preservation of anatomical details is crucial for accurate and safe diagnosis,such as tumor detection in MRI and vascular imaging in CT.展开更多
Predicting human motion based on historical motion sequences is a fundamental problem in computer vision,which is at the core of many applications.Existing approaches primarily focus on encoding spatial dependencies a...Predicting human motion based on historical motion sequences is a fundamental problem in computer vision,which is at the core of many applications.Existing approaches primarily focus on encoding spatial dependencies among human joints while ignoring the temporal cues and the complex relationships across non-consecutive frames.These limitations hinder the model’s ability to generate accurate predictions over longer time horizons and in scenarios with complex motion patterns.To address the above problems,we proposed a novel multi-level spatial and temporal learning model,which consists of a Cross Spatial Dependencies Encoding Module(CSM)and a Dynamic Temporal Connection Encoding Module(DTM).Specifically,the CSM is designed to capture complementary local and global spatial dependent information at both the joint level and the joint pair level.We further present DTM to encode diverse temporal evolution contexts and compress motion features to a deep level,enabling the model to capture both short-term and long-term dependencies efficiently.Extensive experiments conducted on the Human 3.6M and CMU Mocap datasets demonstrate that our model achieves state-of-the-art performance in both short-term and long-term predictions,outperforming existing methods by up to 20.3% in accuracy.Furthermore,ablation studies confirm the significant contributions of the CSM and DTM in enhancing prediction accuracy.展开更多
基金supported by the Research Center for Chengdu History and Chengdu Literature[CLWX24004]the Centre for Southeast Asia Economic and Culture Studies[DNY2415]the Sichuan Landscape and Recreation Research Center[JGYQ2025027].
文摘Chengdu teahouses,as core public spaces in marketplace society,have undergone transformative reconstruction-from“containers of everyday life”to“containers of commercial traffic and digital flows”-during the process of modernization.Employing spatial archaeology as a methodology,combined with fieldwork and analysis of historical documents,this study systematically examines the diachronic evolution of architectural forms,functional orientations,and social networks within Chengdu teahouses.The study reveals the logic of spatial reconstruction under the interplay of multiple forces,including cultural heritage preservation,capital-driven development,and technological intervention.The findings identify three paradigms of spatial transformation in teahouses.First,heritage specimenization,which reinforces the continuity of collective memory through symbolic extraction but risks diminishing the vitality of everyday social interactions.Second,consumption upgrading,which caters to the demands of emerging groups through iterative business models yet necessitates vigilance against spatial differentiation eroding marketplace inclusivity.Third,digital parasitism,which expands communicative dimensions through technological empowerment but confronts the risk of flattening localized knowledge.These paradigms reflect both adaptive responses of traditional spaces to contemporary pressure and the tension of reconstruction imposed by instrumental rationality on marketplace networks.The study demonstrates that spatial transformation in Chengdu teahouses is not unidirectional alienation but rather a multifaceted configuration where the continuity of tradition coexists with innovative practices amid functional diversification.This research advocates for striking a balance between the preservation of traditional spaces and modern renewal and explores organic integration approaches for traditional and modern elements,thereby providing a theoretical framework and practical insights for the transformation of traditional public spaces.
基金Under the auspices of National Natural Science Foundation of China(No.42271206)Guangdong Philosophy and Social Science Foundation(No.GD22XGL08)+1 种基金Basic and Applied Basic Research Foundation of Guangzhou(No.2024A04J4541)the Fundamental Research Funds for the Central Universities(No.2024ZYGXZR025,2024ZYGXZR003)。
文摘The rapid development of digital technologies has driven the emergence and popularization of online-to-offline(O2O)retail,reshaping the retail landscape in urban China.However,spatial distribution characteristics and influencing mechanisms of emerging O2O retail have not been thoroughly investigated in extant studies.Taking the central urban area of Guangzhou as the case,this study utilized multi-source data and machine learning methods to explore the distribution characteristics of O2O retail space and to further identify the nonlinear effects of the built environment,sociodemographic,and economic factors on its distribution.The results revealed that O2O retail space exhibited a‘single-center’distribution pattern,in contrast to the‘multi-center’distribution pattern of traditional retail space.This finding supported the diffusion of innovation hypothesis,highlighting that the expansion of O2O retail modes first spread from traditional developed retail space.Furthermore,spatial heterogeneities were observed across different types of O2O retail space,with O2O in-store showing a‘core-periphery’spatial structure as described by Central Place Theory,whereas O2O delivery displaying a‘horizontal,non-hierarchical,and multi-centered’network structure following Central Flow Theory.Compared to traditional retail space,the distribution of O2O retail space was more influenced by sociodemographic factors such as the proportion of youth,education level,and income level,but less affected by the built environment factors like office and building density.Furthermore,nonlinear effects of these influencing factors on the distribution of O2O retail space were identified,which enriched the existing literature by highlighting effective ranges and threshold effects.These findings provided valuable insights into O2O retail space development in the context of digital transformation.
基金This work is supported by the National Nature Science Foundation of China(NSFC)under Grant No.61571106Jiangsu Natural Science Foundation under Grant No.BK20170757the Natural Science Foundation of the Jiangsu Higher Education Institutions of China under grant No.17KJD510002.
文摘Microphone array-based sound source localization(SSL)is a challenging task in adverse acoustic scenarios.To address this,a novel SSL algorithm based on deep neural network(DNN)using steered response power-phase transform(SRP-PHAT)spatial spectrum as input feature is presented in this paper.Since the SRP-PHAT spatial power spectrum contains spatial location information,it is adopted as the input feature for sound source localization.DNN is exploited to extract the efficient location information from SRP-PHAT spatial power spectrum due to its advantage on extracting high-level features.SRP-PHAT at each steering position within a frame is arranged into a vector,which is treated as DNN input.A DNN model which can map the SRP-PHAT spatial spectrum to the azimuth of sound source is learned from the training signals.The azimuth of sound source is estimated through trained DNN model from the testing signals.Experiment results demonstrate that the proposed algorithm significantly improves localization performance whether the training and testing condition setup are the same or not,and is more robust to noise and reverberation.
基金The Strategic Priority Research Program of the Chinese Academy of Sciences,No.XDA23020101National Natural Science Foundation of China,No.41901181。
文摘The Yangtze River Delta(YRD) is a region in China with a serious contradiction between economic growth and environmental pollution. Exploring the spatiotemporal effects and influencing factors of air pollution in the region is highly important for formulating policies to promote the high-quality development of urban industries. This study uses the spatial Durbin model(SDM) to analyze the local direct and spatial spillover effects of industrial transformation on air pollution and quantifies the contribution of each factor. From 2008 to 2018, there was a significant spatial agglomeration of industrial sulfur dioxide emissions(ISDE) in the YRD, and every 1% increase in ISDE led to a synchronous increase of 0.603% in the ISDE in adjacent cities. The industrial scale index(ISCI) and industrial structure index(ISTI), as the core factors of industrial transformation, significantly affect the emissions of sulfur dioxide in the YRD, and the elastic coefficients are 0.677 and-0.368, respectively. The order of the direct effect of the explanatory variables on local ISDE is ISCI>ISTI>foreign direct investment(FDI)>enterprise technological innovation(ETI)>environmental regulation(ER)> per capita GDP(PGDP). Similarly, the order of the spatial spillover effect of all variables on ISDE in adjacent cities is ISCI>PGDP>FDI>ETI>ISTI>ER, and the coefficients of the ISCI and ISTI are 1.531 and 0.113, respectively. This study contributes to the existing research that verifies the environmental Kuznets curve in the YRD, denies the pollution heaven hypothesis, indicates the Porter hypothesis, and provides empirical evidence for the formation mechanism of regional environmental pollution from a spatial spillover perspective.
基金supported by the National Key Research and Development Program of China(No.2020YFB2010800)the National Natural Science Foundation of China(Nos.51875433 and 92060302)+1 种基金the Natural Science Foundation of Shaanxi Province,China(No.2019KJXX-043,2021JC-04)the Fundamental Research Funds for the Central Universities and the Foundation of Beilin District,China(No.GX2029)。
文摘Blade-health monitoring is intensely required for turbomachinery because of the high failure risk of rotating blades.Blade-Tip Timing(BTT)is considered as the most promising technique for operational blade-vibration monitoring,which obtains the parameters that characterize the blade condition from recorded signals.However,its application is hindered by severe undersampling and stringent probe layouts.An inappropriate probe layout can make most of the existing methods invalid or inaccurate.Additionally,a general conflict arises between the allowed and required layouts because of arrangement restrictions.For the sake of economy and safety,parameter identification based on fewer probes has been preferred by users.In this work,a spatial-transformation-based method for parameter identification is proposed based on a single-probe BTT measurement.To present the general Sampling-Aliasing Frequency(SAFE)map definition,the traditional time-frequency analysis methods are extended to a time-sampling frequency.Then,a SAFE map is projected onto a parameter space using spatial transformation to extract the slope and intercept parameters,which can be physically interpreted as an engine order and a natural frequency using coordinate transformation.Finally,the effectiveness and robustness of the proposed method are verified by simulations and experiments under uniformly and nonuniformly variable speed conditions.
文摘【目的】高分辨率遥感影像语义分割通过精准提取地物信息,为城市规划、土地分析利用提供了重要的数据支持。当前分割方法通常将遥感影像划分为标准块,进行多尺度局部分割和层次推理,未充分考虑影像中的上下文先验知识和局部特征交互能力,影响了推理分割质量。【方法】为了解决这一问题,本文提出了一种联合跨尺度注意力和语义视觉Transformer的遥感影像分割框架(Cross-scale Attention Transformer,CATrans),融合跨尺度注意力模块和语义视觉Transformer,提取上下文先验知识增强局部特征表示和分割性能。首先,跨尺度注意力模块通过空间和通道两个维度进行并行特征处理,分析浅层-深层和局部-全局特征之间的依赖关系,提升对遥感影像中不同粒度对象的注意力。其次,语义视觉Transformer通过空间注意力机制捕捉上下文语义信息,建模语义信息之间的依赖关系。【结果】本文在DeepGlobe、Inria Aerial和LoveDA数据集上进行对比实验,结果表明:CATrans的分割性能优于现有的WSDNet(Discrete Wavelet Smooth Network)和ISDNet(Integrating Shallow and Deep Network)等分割算法,分别取得了76.2%、79.2%、54.2%的平均交并比(Mean Intersection over Union,mIoU)和86.5%、87.8%、66.8%的平均F1得分(Mean F1 Score,mF1),推理速度分别达到38.1 FPS、13.2 FPS和95.22 FPS。相较于本文所对比的最佳方法WSDNet,mIoU和mF1在3个数据集中分别提升2.1%、4.0%、5.3%和1.3%、1.8%、5.6%,在每类地物的分割中都具有显著优势。【结论】本方法实现了高效率、高精度的高分辨率遥感影像语义分割。
文摘Relative to hospitalized patient information, outpatient admission information is relatively simple. It only includes the patient admission time, place of residence and other information. Traditionally, the excavation of this information is not sufficient. However, when a large number of patients admitted time and residence information combined to consider, and add some data mining technology, some of the previously ignored regular information is likely to be found. Using 5 years of data mining research and admission data from a paediatric department at a large women’s and children’s hospital in China, we found important fluctuation rules regarding admissions using wavelet analysis on hospital admission data among different scales of cyclical fluctuations. Method: Seasonal distribution of patient number was analysed based on Haar wavelet transformation, and level 3 and level 2 of wavelets were extracted out to fit the data. The distribution function of hospitalized patients was visualized by kernel density estimation. Using linear regression and ARIMA (autoregressive integrated moving average model) predict the seasonally number of patients in the future. Results: The data analysis demonstrates the total surge of inpatients was decomposed into one mother wavelet and five small wavelets, each of which represents different time frequency. Besides, as distance from hospital increases, the number of patients decreased exponentially. The seasonal factors are the largest time factor influencing the number changes of patients. Conclusion: By wavelet analysis and the improved prediction model, we could make forecast on the future inpatient number trend and prove factors such as geographic position is influential on inpatient amount. Additionally, the concept of data mining based on spatial distribution and spectral analysis could be applied to other aspects of social management.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.61971300,11604292,61905208,11674239,and 11621101)the Scientific and Technological Innovation Programs(STIP)of Higher Education Institutions in Shanxi Province,China(Grant Nos.2019L0159 and 2019L0146)the Postdoctoral Science Foundation of China(Grant Nos.2017T100430 and 2018M632455).
文摘By introducing an optic-null medium into the finite embedded transformation,a reflectionless spatial beam bender is designed,which can steer the output beam by a fixed pre-designed angleβfor an arbitrary incident angle.The bending angleβof the beam bender is determined by the geometrical angle of the device,which can be changed by simply choosing different geometrical angles.For various bending angles,the designed spatial beam bender can be realized by the same materials(i.e.,an optic-null medium),which is a homogenous anisotropic material.Numerical simulations verify the reflectionless bending effect and rotated imaging ability of the proposed beam bender.A reduction model of the optic-null medium is studied,which can also be used for a reflectionless spatial beam bender with a pre-designed bending angle.
基金Supported by National Natural Science Foundation of China(Grant No.51377015)China Post Doctor Project(Grant No.136413)Science&Technology Department of Sichuan Province,China(Grant No.2013HH0059)
文摘Eddy current pulsed thermography（ECPT） is an emerging Non-destructive testing and evaluation（NDT E） technique, which uses hybrid eddy current and thermography NDT E techniques that enhances the detectability from their compensation. Currently, this technique is limited by the manual selection of proper contrast frames and the issue of improving the efficiency of defect detection of complex structure samples remains a challenge. In order to select a specific frame from transient thermal image sequences to maximize the contrast of thermal variation and defect pattern from complex structure samples, an energy driven approach to compute the coefficient energy of wavelet transform is proposed which has the potential of automatically selecting both optimal transient frame and spatial scale for defect detection using ECPT. According to analysis of the variation of different frequency component and the comparison study of the detection performance of different scale and wavelets, the frame at the end of heating phase is automatically selected as an optimal transient frame for defect detection. In addition, the detection capabilities of the complex structure samples can be enhanced through proper spatial scale and wavelet selection. The proposed method has successfully been applied to low speed impact damage detection of carbon fibre reinforced polymer（CFRP） composite as well as providing the guidance to improve the detectability of ECPT technique.
文摘This letter exploits fundamental characteristics of a wavelet transform image to form a progressive octave-based spatial resolution. Each wavelet subband is coded based on zeroblock and quardtree partitioning ordering scheme with memory optimization technique. The method proposed in this letter is of low complexity and efficient for Internet plug-in software.
文摘The spatial transformations can be observed at different religious-historic towns of India due to urbanization. Research is based upon fact that there is substantial change in the built environment because of spatial transformations at the religious-historic towns. The process of modernization in the functions and spatial layout is unavoidable at any historic town. The study attempts to focus on various urban historic conservation components, including the look of historic buildings, their earlier uses, and its immediate surroundings to improve the built environment of historic towns. A theoretical framework for the urban conservation of ancient towns is the main objective of study. How to modernize the historic conservation function while preserving the space’s texture and integrity. The research started with the investigation of the morphological growth of Mathura district, India through satellite images and in-depth study of the evolution process of street network in Vrindavan town, which is one of the main temple towns of Mathura district. There is a significant difference in the layout & architectural character of old part and the newly developed Vrindavan. Due to increased accessibility and movement, the spatial structure of traditional religious precincts, which were once local integration centres, has significantly changed. Increasing & changing mode of transportation and further increase in the religious tourism might be the cause or a big reason for the spatial transformations and correspondingly there is a challenge to conserve & preserve the religious precincts of historic towns. The study tries to analyze spatial transformations with the help of Historical GIS at different scales of urban form. Suggestive measures to conserve the environmental ambience of religious-historic towns are the outcome of the research.
文摘本文提出了一种基于双交叉注意力融合的Swin-AK Transformer(Swin Transformer based on alterable kernel convolution)和手工特征相结合的智能手机拍摄图像质量评价方法。首先,提取了影响图像质量的手工特征,这些特征可以捕捉到图像中细微的视觉变化;其次,提出了Swin-AK Transformer,增强了模型对局部信息的提取和处理能力。此外,本文设计了双交叉注意力融合模块,结合空间注意力和通道注意力机制,融合了手工特征与深度特征,实现了更加精确的图像质量预测。实验结果表明,在SPAQ和LIVE-C数据集上,皮尔森线性相关系数分别达到0.932和0.885,斯皮尔曼等级排序相关系数分别达到0.929和0.858。上述结果证明了本文提出的方法能够有效地预测智能手机拍摄图像的质量。
文摘Background Several face detection and recogni tion methods have been proposed in the past decades that have excellent performance.The conventional face recognition pipeline comprises the following:(1)face detection,(2)face alignment,(3)feature extraction,and(4)similarity,which are independent of each other.The separate facial analysis stages lead to redundant model calculations,and are difficult for use in end-to-end training.Methods In this paper,we propose a novel end-to-end trainable convolutional network framework for face detection and recognition,in which a geometric transformation matrix is directly learned to align the faces rather than predicting the facial landmarks.In the training stage,our single CNN model is supervised only by face bounding boxes and personal identities,which are publicly available from WIDER FACE and CASIA-WebFace datasets.Our model is tested on Face Detection Dataset and Benchmark(FDDB)and Labeled Face in the Wild(LFW)datasets.Results The results show 89.24%recall for face detection tasks and 98.63%accura cy for face recognition tasks.
文摘提出了一种Transformer与图网络相结合的网络模型,用于对视觉传感器采集到的视频图像进行三维人体姿态估计。Transformer能够有效地从二维关键关节点中提取时空维度高相关性特征,而图网络则能够感知细节相关性特征,通过融合这两种网络结构,提高了三维姿态估计的精度。在公开数据集Human3.6M上进行了仿真实验,验证了Transformer与图卷积融合算法的性能。实验结果显示,最终估计得到的三维人体关节点的平均关节点位置偏差(Mean Per Joint Position Error,MPJPE)为38.4 mm,相较于现有方法有一定提升,表明该方法具有较强的应用价值,可应用于许多下游相关工作中。
基金the Foshan Science and technology Innovation Team Project(No.FS0AA-KJ919-4402-0060)the National Natural Science Foundation of China(No.62263018)。
文摘In view of the weak ability of the convolutional neural networks to explicitly learn spatial invariance and the probabilistic loss of discriminative features caused by occlusion and background interference in pedestrian re-identification tasks,a person re-identification method combining spatial feature learning and multi-granularity feature fusion was proposed.First,an attention spatial transformation network(A-STN)is proposed to learn spatial features and solve the problem of misalignment of pedestrian spatial features.Then the network was divided into a global branch,a local coarse-grained fusion branch,and a local fine-grained fusion branch to extract pedestrian global features,coarse-grained fusion features,and fine-grained fusion features,respectively.Among them,the global branch enriches the global features by fusing different pooling features.The local coarse-grained fusion branch uses an overlay pooling to enhance each local feature while learning the correlation relationship between multi-granularity features.The local fine-grained fusion branch uses a differential pooling to obtain the differential features that were fused with global features to learn the relationship between pedestrian local features and pedestrian global features.Finally,the proposed method was compared on three public datasets:Market1501,DukeMTMC-ReID and CUHK03.The experimental results were better than those of the comparative methods,which verifies the effectiveness of the proposed method.
文摘To enable proper diagnosis of a patient,medical images must demonstrate no presence of noise and artifacts.The major hurdle lies in acquiring these images in such a manner that extraneous variables,causing distortions in the form of noise and artifacts,are kept to a bare minimum.The unexpected change realized during the acquisition process specifically attacks the integrity of the image’s quality,while indirectly attacking the effectiveness of the diagnostic process.It is thus crucial that this is attended to with maximum efficiency at the level of pertinent expertise.The solution to these challenges presents a complex dilemma at the acquisition stage,where image processing techniques must be adopted.The necessity of this mandatory image pre-processing step underpins the implementation of traditional state-of-the-art methods to create functional and robust denoising or recovery devices.This article hereby provides an extensive systematic review of the above techniques,with the purpose of presenting a systematic evaluation of their effect on medical images under three different distributions of noise,i.e.,Gaussian,Poisson,and Rician.A thorough analysis of these methods is conducted using eight evaluation parameters to highlight the unique features of each method.The covered denoising methods are essential in actual clinical scenarios where the preservation of anatomical details is crucial for accurate and safe diagnosis,such as tumor detection in MRI and vascular imaging in CT.
基金supported by the Urgent Need for Overseas Talent Project of Jiangxi Province(Grant No.20223BCJ25040)the Thousand Talents Plan of Jiangxi Province(Grant No.jxsg2023101085)+3 种基金the National Natural Science Foundation of China(Grant No.62106093)the Natural Science Foundation of Jiangxi(Grant Nos.20224BAB212011,20232BAB212008,20242BAB25078,and 20232BAB202051)The Youth Talent Cultivation Innovation Fund Project of Nanchang University(Grant No.XX202506030015)funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R759),Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘Predicting human motion based on historical motion sequences is a fundamental problem in computer vision,which is at the core of many applications.Existing approaches primarily focus on encoding spatial dependencies among human joints while ignoring the temporal cues and the complex relationships across non-consecutive frames.These limitations hinder the model’s ability to generate accurate predictions over longer time horizons and in scenarios with complex motion patterns.To address the above problems,we proposed a novel multi-level spatial and temporal learning model,which consists of a Cross Spatial Dependencies Encoding Module(CSM)and a Dynamic Temporal Connection Encoding Module(DTM).Specifically,the CSM is designed to capture complementary local and global spatial dependent information at both the joint level and the joint pair level.We further present DTM to encode diverse temporal evolution contexts and compress motion features to a deep level,enabling the model to capture both short-term and long-term dependencies efficiently.Extensive experiments conducted on the Human 3.6M and CMU Mocap datasets demonstrate that our model achieves state-of-the-art performance in both short-term and long-term predictions,outperforming existing methods by up to 20.3% in accuracy.Furthermore,ablation studies confirm the significant contributions of the CSM and DTM in enhancing prediction accuracy.
