This essay aims at an analysis of Murphy's failure of his quest for freedom from the aspect of narrative temporality whichwill focus on order, duration and frequency. From the perspective of temporal order, the es...This essay aims at an analysis of Murphy's failure of his quest for freedom from the aspect of narrative temporality whichwill focus on order, duration and frequency. From the perspective of temporal order, the essay will endeavor to probe into the pro-lepsis in the novel to analyse predetermined narrative and its function. Considering with temporal duration, this essay will penetratethe deceleration of narrative—scene and pause in Murphy to explore its thematic meaning. To investigate the temporal frequency,iterative narrative within text will be exemplified to discuss closed systems in the novel. Based on the analysis of the narrativity tem-porality in Murphy, the essay will focuses on that how its narrative strategy well embodies its theme—the predetermined failure ofMurphy's quest for freedom in closed systems.展开更多
Pioneering U-Pb isotopic studies by a small group of workers in the mid-late 1980s demonstrated the feasibility of using rare accessory mineral chronometers in mafic(gabbroic)intrusive rocks.These examples showed that...Pioneering U-Pb isotopic studies by a small group of workers in the mid-late 1980s demonstrated the feasibility of using rare accessory mineral chronometers in mafic(gabbroic)intrusive rocks.These examples showed that mafic layered intrusions and diabase/dolerite dyke swarms alike crystallized high-temperature展开更多
Mandarin在(pinyin:zài)is the most frequently used character in representing spatial and temporal relationship.Current studies mostly focus on its lexical meaning and syntactic structure while cognitive features o...Mandarin在(pinyin:zài)is the most frequently used character in representing spatial and temporal relationship.Current studies mostly focus on its lexical meaning and syntactic structure while cognitive features of its grammatical categories have been neglected.This paper investigates into the categorization of zài by conducting a morphosyntactic test among College English majors in China.The results show that:prototypes are organizing the grammatical categories of zài at all levels in terms of intra-categorial gradience;the semantic construal of zài construction could significantly influence the accuracy of the grammatical categorization of zài;the syntactic structure can provide viable cue for the identification of grammatical categories of zài;spatiality,temporality and the status of existing are three essential semantic features encoded by zài,the concurrence of which leads to various degree of inter-categorial vagueness,indicating a conflict between the rigid grammatical classification and the indeterminate nature of the grammatical functions of zai,suggesting the necessity to reconsider the efficacy of applying indiscriminately the Anglo-Saxon grammar into the study of Chinese spatial-temporal constructions.展开更多
OpenStreetMap(OSM)is a dataset in constant change and this dynamic needs to be better understood.Based on 12-year time series of seven OSM data contribution activities extracted from 20 large cities worldwide,we inves...OpenStreetMap(OSM)is a dataset in constant change and this dynamic needs to be better understood.Based on 12-year time series of seven OSM data contribution activities extracted from 20 large cities worldwide,we investigate the temporal dynamic of OSM data production,more specifically,the auto-and cross-correlation,temporal trend,and annual seasonality of these activities.Furthermore,we evaluate and compare nine different temporal regression methods for forecasting such activities in horizons of 1–4 weeks.Several insights could be obtained from our analyses,including that the contribution activities tend to grown linearly in a moderate intra-annual cycle.Also,the performance of the temporal forecasting methods shows that they yield in general more accurate estimations of future contribution activities than a baseline metric,i.e.the arithmetic average of recent previous observations.In particular,the well-known ARIMA and the exponentially weighted moving average methods have shown the best performances.展开更多
Sepsis remains a leading cause morbidity and mortality worldwide;effective targeted therapies remain elusive due to its inherent heterogeneity and dynamic temporal evolution.Existing frameworks often focus on either t...Sepsis remains a leading cause morbidity and mortality worldwide;effective targeted therapies remain elusive due to its inherent heterogeneity and dynamic temporal evolution.Existing frameworks often focus on either the diverse manifestations of sepsis or its progression over time,but fail to integrate these critical aspects.In this review,we propose a novel spatial-temporal framework that integrates both the heterogeneity and temporality of sepsis.The framework consists of two key dimensions:The cross-sectional(heterogeneity)dimension,which addresses pathogen variability,host factors,and pathogen-host interactions;The longitudinal(temporality)dimension,which explores the dynamic evolution of sepsis and the need for adaptive,real-time interventions.Given the complexity of multidimensional temporal data,big data techniques have the potential to integrate these data and decompose sepsis into distinct disease subtypes.Stratification facilitates the development of personalized therapeutic approaches tailored to specific subtypes.Moreover,methods,such as reinforcement learning,can track the dynamic transitions between these subtypes,enabling real-time adaptation of treatment strategies.展开更多
Question answering is an important problem that aims to deliver specific answers to questions posed by humans in natural language.How to efficiently identify the exact answer with respect to a given question has becom...Question answering is an important problem that aims to deliver specific answers to questions posed by humans in natural language.How to efficiently identify the exact answer with respect to a given question has become an active line of research.Previous approaches in factoid question answering tasks typically focus on modeling the semantic relevance or syntactic relationship between a given question and its corresponding answer.Most of these models suffer when a question contains very little content that is indicative of the answer.In this paper,we devise an architecture named the temporality-enhanced knowledge memory network(TE-KMN) and apply the model to a factoid question answering dataset from a trivia competition called quiz bowl.Unlike most of the existing approaches,our model encodes not only the content of questions and answers,but also the temporal cues in a sequence of ordered sentences which gradually remark the answer.Moreover,our model collaboratively uses external knowledge for a better understanding of a given question.The experimental results demonstrate that our method achieves better performance than several state-of-the-art methods.展开更多
The concept of time is encoded in language in different ways,which is reflected in temporal expressions,tenses,aspects and word order in languages.Some scholars claim that Chinese is a language characterized by tempor...The concept of time is encoded in language in different ways,which is reflected in temporal expressions,tenses,aspects and word order in languages.Some scholars claim that Chinese is a language characterized by temporality.This paper aims to explore what the temporality of Chinese refers to in both its narrow and broad senses,and the classification of temporality under the guidance of the iconicity of sequence as well as the word order principles.Moreover,some scholars hold a seemingly opposite view that Chinese is a language of spatiality—whether the two views are contradictory to each other is also discussed in the current study.The research mainly adopts the methods of literature analysis and retrospection.It is found that the temporality of Chinese mainly refers to the correspondence between the sequence of linguistic units and the temporal order of events or states represented in the conceptual world,and that temporality can be classified into natural temporality,cognitive temporality,and functional temporality.Then,this paper concludes that the viewpoint that Chinese is a language of temporality does not conflict with the opinion that Chinese is characterized by spatiality,as these two traits are concerned with two irrelevant aspects of language.However,it will posit that whether the property of temporality is consistent at every linguistic level in Chinese,even including characters,needs further exploration.展开更多
The temporal pole(TP),one of the most expanded cortical regions in humans relative to other primates,plays a crucial role in human language processing.It is also one of the most structurally and functionally asymmetri...The temporal pole(TP),one of the most expanded cortical regions in humans relative to other primates,plays a crucial role in human language processing.It is also one of the most structurally and functionally asymmetric regions.However,whether the functional architecture of the TP is shared by humans and macaques is an open question.We used spectral clustering algorithms to define a cross-species fine-grained TP atlas with different anatomical connectivity patterns.We identified three similar subregions,two ventral and one dorsal,within the TP in both humans and macaques.The parcellation scheme for the TP was validated using functional gradient mapping,anatomical connectivity and resting-state functional connectivity pattern analysis,and functional characterization.Furthermore,in conjunction with the Allen Human Brain Atlas,we revealed the molecular basis for the functional connectivity patterns of each human TP subregion.In addition,we compared the hemispheric asymmetry in mean gray matter volume,anatomical connectivity fingerprints,and whole brain functional connectivity patterns to reveal the evolutionary differences in the TP and found different asymmetric patterns between humans and macaques.In conclusion,our findings reveal that the asymmetry in structure and connectivity may underpin the hemispheric functional specialization of the brain and provide a novel insight into understanding the evolutionary origin of the TP.展开更多
Brain lesions,such as those caused by stroke or traumatic brain injury(TBI),frequently result in persistent motor and cognitive impairments that significantly affect the individual patient's quality of life.Despit...Brain lesions,such as those caused by stroke or traumatic brain injury(TBI),frequently result in persistent motor and cognitive impairments that significantly affect the individual patient's quality of life.Despite differences in the mechanisms of injury,both conditions share a high prevalence of motor and cognitive impairments.These deficits show only limited natural recovery.展开更多
Central Asia is characterized by an arid climate and widespread desert distribution,with its sustainable development severely constrained by dust events.An objective understanding of the spatiotemporal patterns and dr...Central Asia is characterized by an arid climate and widespread desert distribution,with its sustainable development severely constrained by dust events.An objective understanding of the spatiotemporal patterns and driving forces of dust weather is highly important in this area.Based on the meteorological observations from 2000 to 2020,we examined the spatiotemporal characteristics of dust weather in the five Central Asian countries(Kazakhstan,Uzbekistan,Kyrgyzstan,Turkmenistan,and Tajikistan)via Theil-Sen trend analysis and Geodetector modeling method,quantitatively revealing the influence of environmental factors,such as temperature,precipitation,and vegetation,on the frequency of dust weather.The results showed that:(1)dust weather in Central Asia was mainly distributed in a large''dust belt''extending from west to east from northern part of the Caspian lowland desert,and concentrated in basins,plains,and other low-altitude areas.Strong dust weather mainly occurred in northern areas of the Aral Sea and southern edge of Central Asia,with a maximum annual frequency of 21.9%;(2)strong dust weather in Central Asia has fluctuated and slightly decreased since 2001.The highest frequency(1.1%)occurred in spring(from March to June);(3)from 2000 to 2020,changes such as spot shifting and shrinking occurred in the four main source areas(north of the Aral Sea,Kyzylkum Desert,Karakum Desert,and Garabogazköl Bay region),where sandstorms occurred in Central Asia,and northern Caspian lowland desert became the most important low-emission dust source in Central Asia;and(4)the combined effect of soil moisture and air temperature has the most significant influence on dust weather in Central Asia.This study provides a theoretical basis for sand prevention and sand control in Central Asia.In the future,Central Asia should focus on the rational utilization of land and water resources,and implement human interventions such as vegetation restoration and optimization of irrigation methods to curb further desertification in this area.展开更多
Evaluating the effectivenes s of forest restoration projects is crucial for designing adaptive restoration strategies.However,existing studies have primarily focused on ecological outcomes while overlooking cost input...Evaluating the effectivenes s of forest restoration projects is crucial for designing adaptive restoration strategies.However,existing studies have primarily focused on ecological outcomes while overlooking cost inputs.This gap can lead to increased uncertainties in restoration planning.Here we investigated forest dynamics in China's Upper Yangtze River Basin(UYRB)using kernel Normalized Difference Vegetation Index(kNDVI),Leaf Area Index(LAI),Gross Primary Productivity(GPP),Ku-band Vegetation Optical Depth(Ku-VOD)time series and climate data from1982 to 2020.Subsequently,we employed a residual trend analysis integrating temporal effects to determine the relative contributions of climate change and human activities to forest dynamics before and after the implementation of forest restoration engineering in 1998.Additionally,we developed an Afforestation Efficiency Index(AEI)to quantitatively assess the cost efficiency of afforestation projects.Results indicated that forest in the UYRB showed sustained increases during 1982-2020,with most areas experiencing greater growth after 1998 than before.Temporal effects of climatic factors influenced over 42.7%of the forest,and incorporating time-lag and cumulative effects enhanced climate-based explanations of forest variations by 1.61-24.73%.Human activities emerged as the dominant driver of forest dynamics post 1998,whereas climate variables predominated before this period.The cost-effectiveness of forest restoration projects in the UYRB typically ranges from moderate to high,with higher success predominantly observed in the northeastern and eastern counties,while the central,western,and northwestern counties mainly showed relatively low efficiency.These findings stress the need for assessing forest restoration outcomes from both ecological and cost perspectives,and can offer valuable insights for optimizing the layout of forest restoration initiatives in the UYRB.展开更多
Video emotion recognition is widely used due to its alignment with the temporal characteristics of human emotional expression,but existingmodels have significant shortcomings.On the one hand,Transformermultihead self-...Video emotion recognition is widely used due to its alignment with the temporal characteristics of human emotional expression,but existingmodels have significant shortcomings.On the one hand,Transformermultihead self-attention modeling of global temporal dependency has problems of high computational overhead and feature similarity.On the other hand,fixed-size convolution kernels are often used,which have weak perception ability for emotional regions of different scales.Therefore,this paper proposes a video emotion recognition model that combines multi-scale region-aware convolution with temporal interactive sampling.In terms of space,multi-branch large-kernel stripe convolution is used to perceive emotional region features at different scales,and attention weights are generated for each scale feature.In terms of time,multi-layer odd-even down-sampling is performed on the time series,and oddeven sub-sequence interaction is performed to solve the problem of feature similarity,while reducing computational costs due to the linear relationship between sampling and convolution overhead.This paper was tested on CMU-MOSI,CMU-MOSEI,and Hume Reaction.The Acc-2 reached 83.4%,85.2%,and 81.2%,respectively.The experimental results show that the model can significantly improve the accuracy of emotion recognition.展开更多
End-to-end Temporal Action Detection(TAD)has achieved remarkable progress in recent years,driven by innovations in model architectures and the emergence of Video Foundation Models(VFMs).However,existing TAD methods th...End-to-end Temporal Action Detection(TAD)has achieved remarkable progress in recent years,driven by innovations in model architectures and the emergence of Video Foundation Models(VFMs).However,existing TAD methods that perform full fine-tuning of pretrained video models often incur substantial computational costs,which become particularly pronounced when processing long video sequences.Moreover,the need for precise temporal boundary annotations makes data labeling extremely expensive.In low-resource settings where annotated samples are scarce,direct fine-tuning tends to cause overfitting.To address these challenges,we introduce Dynamic LowRank Adapter(DyLoRA),a lightweight fine-tuning framework tailored specifically for the TAD task.Built upon the Low-Rank Adaptation(LoRA)architecture,DyLoRA adapts only the key layers of the pretrained model via low-rank decomposition,reducing the number of trainable parameters to less than 5%of full fine-tuning methods.This significantly lowers memory consumption and mitigates overfitting in low-resource settings.Notably,DyLoRA enhances the temporal modeling capability of pretrained models by optimizing temporal dimension weights,thereby alleviating the representation misalignment of temporal features.Experimental results demonstrate that DyLoRA-TAD achieves impressive performance,with 73.9%mAP on THUMOS14,39.52%on ActivityNet-1.3,and 28.2%on Charades,substantially surpassing the best traditional feature-based methods.展开更多
Cerebral ischemia restricts cerebral blood flow(CBF),leading to unstable hemodynamics.Past studies of ischemia mainly focused on cortical CBF reduction.However,its impact on hemodynamic changes,especially temporal var...Cerebral ischemia restricts cerebral blood flow(CBF),leading to unstable hemodynamics.Past studies of ischemia mainly focused on cortical CBF reduction.However,its impact on hemodynamic changes,especially temporal varying characteristics,remains poorly understood.Here,we collected cortical resting-state CBF in rats with left carotid artery blockage during occlusion–reperfusion,and measured the temporal variability and changes in laterality using a novel state-space method.This method was also applied to stroke EEG datasets to validate its effectiveness.After arterial occlusion,the left marginal motor,sensory,auditory,and visual cortices exhibited severe temporal variability impairments.The laterality analysis indicated that affected left regions showed inferior unilateral mean,inter-hemispheric transition probability,time fraction,and laterality duration,while the right side had a higher laterality time fraction and duration.These impairments recovered partially following blood flow restoration.Besides,the ischemic state-space metrics were positively correlated with the pre-occlusion baseline appearance.Stroke patients exhibited impaired temporal variability in the affected ischemic hemisphere.The state-space analysis revealed damaged CBF temporal variability during cerebral ischemia and predicted baseline-ischemia connections.展开更多
Accurate precipitation estimation in semiarid,topographically complicated areas is critical for water resource management and climate risk monitoring.This work provides a detailed,multi-scale evaluation of four major ...Accurate precipitation estimation in semiarid,topographically complicated areas is critical for water resource management and climate risk monitoring.This work provides a detailed,multi-scale evaluation of four major satellite precipitation products(CHIRPS,PERSIANN-CDR,IMERG-F v07,and GSMaP)over Isfahan province,Iran,over a 9-year period(2015-2023).The performance of these products was benchmarked against a dense network of 98 rain gauges using a suite of continuous and categorical statistical metrics,following a two-stage quality control protocol to remove outliers and false alarms.The results revealed that the performance of all products improves with temporal aggregation.At the daily level,GSMaP performed marginally better,although all products were linked with considerable uncertainty.At the monthly and annual levels,the GPM-era products(IMERG and GSMaP)clearly beat the other two,establishing themselves as dependable tools for long-term hydro-climatological studies.Error analysis revealed that topography is the dominant regulating factor,creating a systematic elevationdependent bias,largely characterized by underestimation from most products in high-elevation areas,though the PERSIANN-CDR product exhibited a contrasting overestimation tendency.Finally,the findings highlight the importance of implementing local,elevation-dependent calibration before deploying these products in hydrological modeling.展开更多
LiDAR and camera are two of the most common sensors used in the fields of robot perception,autonomous driving,augmented reality,and virtual reality,where these sensors are widely used to perform various tasks such as ...LiDAR and camera are two of the most common sensors used in the fields of robot perception,autonomous driving,augmented reality,and virtual reality,where these sensors are widely used to perform various tasks such as odometry estimation and 3D reconstruction.Fusing the information from these two sensors can significantly increase the robustness and accuracy of these perception tasks.The extrinsic calibration between cameras and LiDAR is a fundamental prerequisite for multimodal systems.Recently,extensive studies have been conducted on the calibration of extrinsic parameters.Although several calibration methods facilitate sensor fusion,a comprehensive summary for researchers and,especially,non-expert users is lacking.Thus,we present an overview of extrinsic calibration and discuss diverse calibration methods from the perspective of calibration system design.Based on the calibration information sources,this study classifies these methods as target-based or targetless.For each type of calibration method,further classification was performed according to the diverse types of features or constraints used in the calibration process,and their detailed implementations and key characteristics were introduced.Thereafter,calibration-accuracy evaluation methods are presented.Finally,we comprehensively compare the advantages and disadvantages of each calibration method and suggest directions for practical applications and future research.展开更多
This paper investigates the challenges of structural inconsistency,matching accuracy degradation,and trajectory interruptions caused by high-speed motion,frequent occlusions,and appearance variations of unmanned aeria...This paper investigates the challenges of structural inconsistency,matching accuracy degradation,and trajectory interruptions caused by high-speed motion,frequent occlusions,and appearance variations of unmanned aerial vehicle(UAV) targets in low-altitude airspace.A novel UAV visual tracking method is proposed for dynamic structural distortions,with a focus on structural consistency modeling to improve system robustness in complex scenarios.Unlike prior methods such as STARK,which rely on spatio-temporal prediction,and KeepTrack,which emphasize template maintenance,our approach enforces structural-level consistency between historical and current features,thereby addressing UAV-specific issues of rapid maneuvering and environmental complexity.The proposed framework features a structure-aware architecture that incorporates dual complementary mechanisms serving as spatial completion and temporal restoration components.First,a multi-scale structure extraction module with adaptive anchor scheduling is developed to dynamically perceive spatial target shape and generate high-quality proposals.Second,a structural memory module is designed to reconstruct local regions by leveraging high-confidence historical structural representations,thereby maintaining spatiotemporal coherence across frames.Furthermore,a structural verification mechanism coupled with a meta-learning-driven re-identification strategy is introduced to detect abrupt structural distortions and adaptively update templates,significantly improving resilience against disturbances.Overall,the main contributions of this paper can be summarized as follows:(1) introducing structural consistency modeling into UAV visual tracking for the first time;(2) designing a unified framework that combines adaptive proposal generation,full-image matching,and re-identification under structural constraints;and(3) achieving state-of-the-art performance on the anti-UAV benchmark,highlighting the method's practical value in real-world UAV surveillance applications.展开更多
Multivariate time series forecasting plays a crucial role in decision-making for systems like energy grids and transportation networks,where temporal patterns emerge across diverse scales from short-term fluctuations ...Multivariate time series forecasting plays a crucial role in decision-making for systems like energy grids and transportation networks,where temporal patterns emerge across diverse scales from short-term fluctuations to long-term trends.However,existing Transformer-based methods often process data at a single resolution or handle multiple scales independently,overlooking critical cross-scale interactions that influence prediction accuracy.To address this gap,we introduce the Hierarchical Attention Transformer(HAT),which enables direct information exchange between temporal hierarchies through a novel cross-scale attention mechanism.HAT extracts multi-scale features using hierarchical convolutional-recurrent blocks,fuses them via temperature-controlled mechanisms,and optimizes gradient flow with residual connections for stable training.Evaluations on eight benchmark datasets show HAT outperforming state-of-the-art baselines,with average reductions of 8.2%in MSE and 7.5%in MAE across horizons,while achieving a 6.1×training speedup over patch-based methods.These advancements highlight HAT’s potential for applications requiring multi-resolution temporal modeling.展开更多
文摘This essay aims at an analysis of Murphy's failure of his quest for freedom from the aspect of narrative temporality whichwill focus on order, duration and frequency. From the perspective of temporal order, the essay will endeavor to probe into the pro-lepsis in the novel to analyse predetermined narrative and its function. Considering with temporal duration, this essay will penetratethe deceleration of narrative—scene and pause in Murphy to explore its thematic meaning. To investigate the temporal frequency,iterative narrative within text will be exemplified to discuss closed systems in the novel. Based on the analysis of the narrativity tem-porality in Murphy, the essay will focuses on that how its narrative strategy well embodies its theme—the predetermined failure ofMurphy's quest for freedom in closed systems.
文摘Pioneering U-Pb isotopic studies by a small group of workers in the mid-late 1980s demonstrated the feasibility of using rare accessory mineral chronometers in mafic(gabbroic)intrusive rocks.These examples showed that mafic layered intrusions and diabase/dolerite dyke swarms alike crystallized high-temperature
文摘Mandarin在(pinyin:zài)is the most frequently used character in representing spatial and temporal relationship.Current studies mostly focus on its lexical meaning and syntactic structure while cognitive features of its grammatical categories have been neglected.This paper investigates into the categorization of zài by conducting a morphosyntactic test among College English majors in China.The results show that:prototypes are organizing the grammatical categories of zài at all levels in terms of intra-categorial gradience;the semantic construal of zài construction could significantly influence the accuracy of the grammatical categorization of zài;the syntactic structure can provide viable cue for the identification of grammatical categories of zài;spatiality,temporality and the status of existing are three essential semantic features encoded by zài,the concurrence of which leads to various degree of inter-categorial vagueness,indicating a conflict between the rigid grammatical classification and the indeterminate nature of the grammatical functions of zai,suggesting the necessity to reconsider the efficacy of applying indiscriminately the Anglo-Saxon grammar into the study of Chinese spatial-temporal constructions.
文摘OpenStreetMap(OSM)is a dataset in constant change and this dynamic needs to be better understood.Based on 12-year time series of seven OSM data contribution activities extracted from 20 large cities worldwide,we investigate the temporal dynamic of OSM data production,more specifically,the auto-and cross-correlation,temporal trend,and annual seasonality of these activities.Furthermore,we evaluate and compare nine different temporal regression methods for forecasting such activities in horizons of 1–4 weeks.Several insights could be obtained from our analyses,including that the contribution activities tend to grown linearly in a moderate intra-annual cycle.Also,the performance of the temporal forecasting methods shows that they yield in general more accurate estimations of future contribution activities than a baseline metric,i.e.the arithmetic average of recent previous observations.In particular,the well-known ARIMA and the exponentially weighted moving average methods have shown the best performances.
基金supported by grants from the Beijing Municipal Natural Science Foundation(No.L222019)CAMS Innovation Fund for Medical Sciences(CIFMS)(No.2024-I2M-C&T-C-002)+1 种基金National High Level Hospital Clinical Research Funding(Nos.2022-PUMCH-B-115 and 2022-PUMCH-D-005)National Key Research and Development Program of China(No.2024YFF1207104).
文摘Sepsis remains a leading cause morbidity and mortality worldwide;effective targeted therapies remain elusive due to its inherent heterogeneity and dynamic temporal evolution.Existing frameworks often focus on either the diverse manifestations of sepsis or its progression over time,but fail to integrate these critical aspects.In this review,we propose a novel spatial-temporal framework that integrates both the heterogeneity and temporality of sepsis.The framework consists of two key dimensions:The cross-sectional(heterogeneity)dimension,which addresses pathogen variability,host factors,and pathogen-host interactions;The longitudinal(temporality)dimension,which explores the dynamic evolution of sepsis and the need for adaptive,real-time interventions.Given the complexity of multidimensional temporal data,big data techniques have the potential to integrate these data and decompose sepsis into distinct disease subtypes.Stratification facilitates the development of personalized therapeutic approaches tailored to specific subtypes.Moreover,methods,such as reinforcement learning,can track the dynamic transitions between these subtypes,enabling real-time adaptation of treatment strategies.
基金supported by the National Basic Research Program(973)of China(No.2015CB352302)the National Natural Science Foundation of China(Nos.61625107,U1611461,U1509206,and 61402403)+2 种基金the Key Program of Zhejiang Province,China(No.2015C01027)the Chinese Knowledge Center for Engineering Sciences and Technologythe Fundamental Research Funds for the Central Universities,China
文摘Question answering is an important problem that aims to deliver specific answers to questions posed by humans in natural language.How to efficiently identify the exact answer with respect to a given question has become an active line of research.Previous approaches in factoid question answering tasks typically focus on modeling the semantic relevance or syntactic relationship between a given question and its corresponding answer.Most of these models suffer when a question contains very little content that is indicative of the answer.In this paper,we devise an architecture named the temporality-enhanced knowledge memory network(TE-KMN) and apply the model to a factoid question answering dataset from a trivia competition called quiz bowl.Unlike most of the existing approaches,our model encodes not only the content of questions and answers,but also the temporal cues in a sequence of ordered sentences which gradually remark the answer.Moreover,our model collaboratively uses external knowledge for a better understanding of a given question.The experimental results demonstrate that our method achieves better performance than several state-of-the-art methods.
文摘The concept of time is encoded in language in different ways,which is reflected in temporal expressions,tenses,aspects and word order in languages.Some scholars claim that Chinese is a language characterized by temporality.This paper aims to explore what the temporality of Chinese refers to in both its narrow and broad senses,and the classification of temporality under the guidance of the iconicity of sequence as well as the word order principles.Moreover,some scholars hold a seemingly opposite view that Chinese is a language of spatiality—whether the two views are contradictory to each other is also discussed in the current study.The research mainly adopts the methods of literature analysis and retrospection.It is found that the temporality of Chinese mainly refers to the correspondence between the sequence of linguistic units and the temporal order of events or states represented in the conceptual world,and that temporality can be classified into natural temporality,cognitive temporality,and functional temporality.Then,this paper concludes that the viewpoint that Chinese is a language of temporality does not conflict with the opinion that Chinese is characterized by spatiality,as these two traits are concerned with two irrelevant aspects of language.However,it will posit that whether the property of temporality is consistent at every linguistic level in Chinese,even including characters,needs further exploration.
基金supported by the Yunnan Fundamental Research Projects(202501AV070005 and 202201BE070001-004).
文摘The temporal pole(TP),one of the most expanded cortical regions in humans relative to other primates,plays a crucial role in human language processing.It is also one of the most structurally and functionally asymmetric regions.However,whether the functional architecture of the TP is shared by humans and macaques is an open question.We used spectral clustering algorithms to define a cross-species fine-grained TP atlas with different anatomical connectivity patterns.We identified three similar subregions,two ventral and one dorsal,within the TP in both humans and macaques.The parcellation scheme for the TP was validated using functional gradient mapping,anatomical connectivity and resting-state functional connectivity pattern analysis,and functional characterization.Furthermore,in conjunction with the Allen Human Brain Atlas,we revealed the molecular basis for the functional connectivity patterns of each human TP subregion.In addition,we compared the hemispheric asymmetry in mean gray matter volume,anatomical connectivity fingerprints,and whole brain functional connectivity patterns to reveal the evolutionary differences in the TP and found different asymmetric patterns between humans and macaques.In conclusion,our findings reveal that the asymmetry in structure and connectivity may underpin the hemispheric functional specialization of the brain and provide a novel insight into understanding the evolutionary origin of the TP.
基金supported by the Defitech Foundation(Morges,CH)to FCHthe Bertarelli Foundation-Catalyst program(Gstaad,CH)to FCH+2 种基金the Wyss Center for Bio and Neuroengineering the Lighthouse Partnership for AI-guided Neuromodulation to FCHthe Fonds de recherche du Quebec-Sante(FRQS#342969)to CEPthe Neuro X Postdoctoral Fellowship Program to CEP。
文摘Brain lesions,such as those caused by stroke or traumatic brain injury(TBI),frequently result in persistent motor and cognitive impairments that significantly affect the individual patient's quality of life.Despite differences in the mechanisms of injury,both conditions share a high prevalence of motor and cognitive impairments.These deficits show only limited natural recovery.
基金funded by the National Natural Science Foundation of China(42571311).
文摘Central Asia is characterized by an arid climate and widespread desert distribution,with its sustainable development severely constrained by dust events.An objective understanding of the spatiotemporal patterns and driving forces of dust weather is highly important in this area.Based on the meteorological observations from 2000 to 2020,we examined the spatiotemporal characteristics of dust weather in the five Central Asian countries(Kazakhstan,Uzbekistan,Kyrgyzstan,Turkmenistan,and Tajikistan)via Theil-Sen trend analysis and Geodetector modeling method,quantitatively revealing the influence of environmental factors,such as temperature,precipitation,and vegetation,on the frequency of dust weather.The results showed that:(1)dust weather in Central Asia was mainly distributed in a large''dust belt''extending from west to east from northern part of the Caspian lowland desert,and concentrated in basins,plains,and other low-altitude areas.Strong dust weather mainly occurred in northern areas of the Aral Sea and southern edge of Central Asia,with a maximum annual frequency of 21.9%;(2)strong dust weather in Central Asia has fluctuated and slightly decreased since 2001.The highest frequency(1.1%)occurred in spring(from March to June);(3)from 2000 to 2020,changes such as spot shifting and shrinking occurred in the four main source areas(north of the Aral Sea,Kyzylkum Desert,Karakum Desert,and Garabogazköl Bay region),where sandstorms occurred in Central Asia,and northern Caspian lowland desert became the most important low-emission dust source in Central Asia;and(4)the combined effect of soil moisture and air temperature has the most significant influence on dust weather in Central Asia.This study provides a theoretical basis for sand prevention and sand control in Central Asia.In the future,Central Asia should focus on the rational utilization of land and water resources,and implement human interventions such as vegetation restoration and optimization of irrigation methods to curb further desertification in this area.
基金supported by the National Natural Science Foundation of China(42071238)the Jiuzhaigou Post-Disaster Restoration and Reconstruction Program(5132202020000046)+1 种基金the Second Tibetan Plateau Scientific Expedition and Research Program(STEP)the Ministry of Science and Technology of the People's Republic of China(2019QZKK0402)。
文摘Evaluating the effectivenes s of forest restoration projects is crucial for designing adaptive restoration strategies.However,existing studies have primarily focused on ecological outcomes while overlooking cost inputs.This gap can lead to increased uncertainties in restoration planning.Here we investigated forest dynamics in China's Upper Yangtze River Basin(UYRB)using kernel Normalized Difference Vegetation Index(kNDVI),Leaf Area Index(LAI),Gross Primary Productivity(GPP),Ku-band Vegetation Optical Depth(Ku-VOD)time series and climate data from1982 to 2020.Subsequently,we employed a residual trend analysis integrating temporal effects to determine the relative contributions of climate change and human activities to forest dynamics before and after the implementation of forest restoration engineering in 1998.Additionally,we developed an Afforestation Efficiency Index(AEI)to quantitatively assess the cost efficiency of afforestation projects.Results indicated that forest in the UYRB showed sustained increases during 1982-2020,with most areas experiencing greater growth after 1998 than before.Temporal effects of climatic factors influenced over 42.7%of the forest,and incorporating time-lag and cumulative effects enhanced climate-based explanations of forest variations by 1.61-24.73%.Human activities emerged as the dominant driver of forest dynamics post 1998,whereas climate variables predominated before this period.The cost-effectiveness of forest restoration projects in the UYRB typically ranges from moderate to high,with higher success predominantly observed in the northeastern and eastern counties,while the central,western,and northwestern counties mainly showed relatively low efficiency.These findings stress the need for assessing forest restoration outcomes from both ecological and cost perspectives,and can offer valuable insights for optimizing the layout of forest restoration initiatives in the UYRB.
基金supported,in part,by the National Nature Science Foundation of China under Grant 62272236,62376128in part,by the Natural Science Foundation of Jiangsu Province under Grant BK20201136,BK20191401.
文摘Video emotion recognition is widely used due to its alignment with the temporal characteristics of human emotional expression,but existingmodels have significant shortcomings.On the one hand,Transformermultihead self-attention modeling of global temporal dependency has problems of high computational overhead and feature similarity.On the other hand,fixed-size convolution kernels are often used,which have weak perception ability for emotional regions of different scales.Therefore,this paper proposes a video emotion recognition model that combines multi-scale region-aware convolution with temporal interactive sampling.In terms of space,multi-branch large-kernel stripe convolution is used to perceive emotional region features at different scales,and attention weights are generated for each scale feature.In terms of time,multi-layer odd-even down-sampling is performed on the time series,and oddeven sub-sequence interaction is performed to solve the problem of feature similarity,while reducing computational costs due to the linear relationship between sampling and convolution overhead.This paper was tested on CMU-MOSI,CMU-MOSEI,and Hume Reaction.The Acc-2 reached 83.4%,85.2%,and 81.2%,respectively.The experimental results show that the model can significantly improve the accuracy of emotion recognition.
基金supported by the National Natural Science Foundation of China(Grant No.62266054)the Major Science and Technology Project of Yunnan Province(Grant No.202402AD080002)the Scientific Research Fund of the Yunnan Provincial Department of Education(Grant No.2025Y0302).
文摘End-to-end Temporal Action Detection(TAD)has achieved remarkable progress in recent years,driven by innovations in model architectures and the emergence of Video Foundation Models(VFMs).However,existing TAD methods that perform full fine-tuning of pretrained video models often incur substantial computational costs,which become particularly pronounced when processing long video sequences.Moreover,the need for precise temporal boundary annotations makes data labeling extremely expensive.In low-resource settings where annotated samples are scarce,direct fine-tuning tends to cause overfitting.To address these challenges,we introduce Dynamic LowRank Adapter(DyLoRA),a lightweight fine-tuning framework tailored specifically for the TAD task.Built upon the Low-Rank Adaptation(LoRA)architecture,DyLoRA adapts only the key layers of the pretrained model via low-rank decomposition,reducing the number of trainable parameters to less than 5%of full fine-tuning methods.This significantly lowers memory consumption and mitigates overfitting in low-resource settings.Notably,DyLoRA enhances the temporal modeling capability of pretrained models by optimizing temporal dimension weights,thereby alleviating the representation misalignment of temporal features.Experimental results demonstrate that DyLoRA-TAD achieves impressive performance,with 73.9%mAP on THUMOS14,39.52%on ActivityNet-1.3,and 28.2%on Charades,substantially surpassing the best traditional feature-based methods.
基金supported by the National Natural Science Foundation of China(82250410380 and 62171101)the Natural Science Foundation of Sichuan Province(24NSFSC6257)the China MOST2030 Brain Project(2022ZD0208500).
文摘Cerebral ischemia restricts cerebral blood flow(CBF),leading to unstable hemodynamics.Past studies of ischemia mainly focused on cortical CBF reduction.However,its impact on hemodynamic changes,especially temporal varying characteristics,remains poorly understood.Here,we collected cortical resting-state CBF in rats with left carotid artery blockage during occlusion–reperfusion,and measured the temporal variability and changes in laterality using a novel state-space method.This method was also applied to stroke EEG datasets to validate its effectiveness.After arterial occlusion,the left marginal motor,sensory,auditory,and visual cortices exhibited severe temporal variability impairments.The laterality analysis indicated that affected left regions showed inferior unilateral mean,inter-hemispheric transition probability,time fraction,and laterality duration,while the right side had a higher laterality time fraction and duration.These impairments recovered partially following blood flow restoration.Besides,the ischemic state-space metrics were positively correlated with the pre-occlusion baseline appearance.Stroke patients exhibited impaired temporal variability in the affected ischemic hemisphere.The state-space analysis revealed damaged CBF temporal variability during cerebral ischemia and predicted baseline-ischemia connections.
文摘Accurate precipitation estimation in semiarid,topographically complicated areas is critical for water resource management and climate risk monitoring.This work provides a detailed,multi-scale evaluation of four major satellite precipitation products(CHIRPS,PERSIANN-CDR,IMERG-F v07,and GSMaP)over Isfahan province,Iran,over a 9-year period(2015-2023).The performance of these products was benchmarked against a dense network of 98 rain gauges using a suite of continuous and categorical statistical metrics,following a two-stage quality control protocol to remove outliers and false alarms.The results revealed that the performance of all products improves with temporal aggregation.At the daily level,GSMaP performed marginally better,although all products were linked with considerable uncertainty.At the monthly and annual levels,the GPM-era products(IMERG and GSMaP)clearly beat the other two,establishing themselves as dependable tools for long-term hydro-climatological studies.Error analysis revealed that topography is the dominant regulating factor,creating a systematic elevationdependent bias,largely characterized by underestimation from most products in high-elevation areas,though the PERSIANN-CDR product exhibited a contrasting overestimation tendency.Finally,the findings highlight the importance of implementing local,elevation-dependent calibration before deploying these products in hydrological modeling.
基金Supported by Beijing Natural Science Foundation(Grant No.L241012)the National Natural Science Foundation of China(Grant No.62572468).
文摘LiDAR and camera are two of the most common sensors used in the fields of robot perception,autonomous driving,augmented reality,and virtual reality,where these sensors are widely used to perform various tasks such as odometry estimation and 3D reconstruction.Fusing the information from these two sensors can significantly increase the robustness and accuracy of these perception tasks.The extrinsic calibration between cameras and LiDAR is a fundamental prerequisite for multimodal systems.Recently,extensive studies have been conducted on the calibration of extrinsic parameters.Although several calibration methods facilitate sensor fusion,a comprehensive summary for researchers and,especially,non-expert users is lacking.Thus,we present an overview of extrinsic calibration and discuss diverse calibration methods from the perspective of calibration system design.Based on the calibration information sources,this study classifies these methods as target-based or targetless.For each type of calibration method,further classification was performed according to the diverse types of features or constraints used in the calibration process,and their detailed implementations and key characteristics were introduced.Thereafter,calibration-accuracy evaluation methods are presented.Finally,we comprehensively compare the advantages and disadvantages of each calibration method and suggest directions for practical applications and future research.
基金Supported by the National Science Foundation of China (No.62571164)the Natural Science Foundation of Heilongjiang Province (No.PL2024F025)the Fundamental Scientific Research Funds of Heilongjiang Province (No.2022-KYYWF-1050)。
文摘This paper investigates the challenges of structural inconsistency,matching accuracy degradation,and trajectory interruptions caused by high-speed motion,frequent occlusions,and appearance variations of unmanned aerial vehicle(UAV) targets in low-altitude airspace.A novel UAV visual tracking method is proposed for dynamic structural distortions,with a focus on structural consistency modeling to improve system robustness in complex scenarios.Unlike prior methods such as STARK,which rely on spatio-temporal prediction,and KeepTrack,which emphasize template maintenance,our approach enforces structural-level consistency between historical and current features,thereby addressing UAV-specific issues of rapid maneuvering and environmental complexity.The proposed framework features a structure-aware architecture that incorporates dual complementary mechanisms serving as spatial completion and temporal restoration components.First,a multi-scale structure extraction module with adaptive anchor scheduling is developed to dynamically perceive spatial target shape and generate high-quality proposals.Second,a structural memory module is designed to reconstruct local regions by leveraging high-confidence historical structural representations,thereby maintaining spatiotemporal coherence across frames.Furthermore,a structural verification mechanism coupled with a meta-learning-driven re-identification strategy is introduced to detect abrupt structural distortions and adaptively update templates,significantly improving resilience against disturbances.Overall,the main contributions of this paper can be summarized as follows:(1) introducing structural consistency modeling into UAV visual tracking for the first time;(2) designing a unified framework that combines adaptive proposal generation,full-image matching,and re-identification under structural constraints;and(3) achieving state-of-the-art performance on the anti-UAV benchmark,highlighting the method's practical value in real-world UAV surveillance applications.
文摘Multivariate time series forecasting plays a crucial role in decision-making for systems like energy grids and transportation networks,where temporal patterns emerge across diverse scales from short-term fluctuations to long-term trends.However,existing Transformer-based methods often process data at a single resolution or handle multiple scales independently,overlooking critical cross-scale interactions that influence prediction accuracy.To address this gap,we introduce the Hierarchical Attention Transformer(HAT),which enables direct information exchange between temporal hierarchies through a novel cross-scale attention mechanism.HAT extracts multi-scale features using hierarchical convolutional-recurrent blocks,fuses them via temperature-controlled mechanisms,and optimizes gradient flow with residual connections for stable training.Evaluations on eight benchmark datasets show HAT outperforming state-of-the-art baselines,with average reductions of 8.2%in MSE and 7.5%in MAE across horizons,while achieving a 6.1×training speedup over patch-based methods.These advancements highlight HAT’s potential for applications requiring multi-resolution temporal modeling.
