Research on human motion prediction has made significant progress due to its importance in the development of various artificial intelligence applications.However,effectively capturing spatio-temporal features for smo...Research on human motion prediction has made significant progress due to its importance in the development of various artificial intelligence applications.However,effectively capturing spatio-temporal features for smoother and more precise human motion prediction remains a challenge.To address these issues,a robust human motion prediction method via integration of spatial and temporal cues(RISTC)has been proposed.This method captures sufficient spatio-temporal correlation of the observable sequence of human poses by utilizing the spatio-temporal mixed feature extractor(MFE).In multi-layer MFEs,the channel-graph united attention blocks extract the augmented spatial features of the human poses in the channel and spatial dimension.Additionally,multi-scale temporal blocks have been designed to effectively capture complicated and highly dynamic temporal information.Our experiments on the Human3.6M and Carnegie Mellon University motion capture(CMU Mocap)datasets show that the proposed network yields higher prediction accuracy than the state-of-the-art methods.展开更多
This study examined the influence of the built environment surrounding rail stations on rail transit ridership and its spatiotemporal variations,aiming to enhance rail transit operational efficiency and inform station...This study examined the influence of the built environment surrounding rail stations on rail transit ridership and its spatiotemporal variations,aiming to enhance rail transit operational efficiency and inform station planning and development.Data from 159 metro stations in Nanjing,collected over a 14-d period,were analyzed to identify changes in weekday and weekend ridership patterns.The analysis included explanatory variables grouped into three categories:urban spatial variables,socioeconomic vari-ables,and transit service variables.A geographically and temporally weighted regression(GTWR)model was developed,and its performance was compared with that of ordinary least squares(OLS)and geographically weighted regression(GWR)models.The results demonstrated that the GTWR model outperformed others in analyzing the relationship between rail transit ridership and the built environment.In addition,the coefficients of explanatory variables showed significant variation across spatiotemporal dimensions,revealing distinct patterns.Notably,the influence of commuter flows led to more pronounced temporal heterogeneity in the coefficients observed on weekdays.These findings offer valuable insights for optimizing urban public transportation systems and advancing integrated urban rail development.展开更多
Action recognition,a fundamental task in the field of video understanding,has been extensively researched and applied.In contrast to an image,a video introduces an extra temporal dimension.However,many existing action...Action recognition,a fundamental task in the field of video understanding,has been extensively researched and applied.In contrast to an image,a video introduces an extra temporal dimension.However,many existing action recognition networks either perform simple temporal fusion through averaging or rely on pre-trained models from image recognition,resulting in limited temporal information extraction capabilities.This work proposes a highly efficient temporal decoding module that can be seamlessly integrated into any action recognition backbone network to enhance the focus on temporal relationships between video frames.Firstly,the decoder initializes a set of learnable queries,termed video-level action category prediction queries.Then,they are combined with the video frame features extracted by the backbone network after self-attention learning to extract video context information.Finally,these prediction queries with rich temporal features are used for category prediction.Experimental results on HMDB51,MSRDailyAct3D,Diving48 and Breakfast datasets show that using TokShift-Transformer and VideoMAE as encoders results in a significant improvement in Top-1 accuracy compared to the original models(TokShift-Transformer and VideoMAE),after introducing the proposed temporal decoder.The introduction of the temporal decoder results in an average performance increase exceeding 11%for TokShift-Transformer and nearly 5%for VideoMAE across the four datasets.Furthermore,the work explores the combination of the decoder with various action recognition networks,including Timesformer,as encoders.This results in an average accuracy improvement of more than 3.5%on the HMDB51 dataset.The code is available at https://github.com/huangturbo/TempDecoder.展开更多
The online version of the original article can be found at:https://www.sciopen.com/article/10.26599/JOTO.2025.9540018 Erratum to Journal of Otology,2025,20(2):123-126.https://doi.org/10.26599/JOTO.2025.9540018 The sur...The online version of the original article can be found at:https://www.sciopen.com/article/10.26599/JOTO.2025.9540018 Erratum to Journal of Otology,2025,20(2):123-126.https://doi.org/10.26599/JOTO.2025.9540018 The surnames and given names of these authors are reversed:Saro-Buendía Miguel,Andresen-Lorca Belén,Pérez-García Alberto,Llópez Carratala Nacho,Carreres Polo Joan,Armengot Carceller Miguel,Perolada Valmaña Jose María.It should be Miguel Saro-Buendía,Belén Andresen-Lorca,Alberto Pérez-García,Nacho Llópez Carratala,Joan Carreres Polo,Miguel Armengot Carceller,Jose María Perolada Valmaña.展开更多
Osteoradionecrosis of the temporal bone(ORN-TB)is usually controlled with conservative measures.However,a temporal bone resection may be required in unresponsive cases.The reconstruction of the resulting defects may b...Osteoradionecrosis of the temporal bone(ORN-TB)is usually controlled with conservative measures.However,a temporal bone resection may be required in unresponsive cases.The reconstruction of the resulting defects may be challenging because of the radiation damage to regional tissues.As a result,distant free flaps may be an optimal choice.For instance,the gracilis muscular free flap(GMFF)has consistent vascular anatomy and can be used to reconstruct small defects.We report three cases of uncontrolled ORN-TB requiring an extensive temporal bone resection followed by vascularized obliteration with a GMFF.The patients reported complete control of the main otologic symptoms(otorrhea,otalgia,and aural fullness)and optimal functional and aesthetic outcomes.Finally,the patients reported significant improvement in quality of life despite early postoperative complications.To our knowledge,the GMFF had not been used to obliterate temporal bone defects in patients with ORN-TB.展开更多
Temporal optics,which enables lossless manipulation of ultrafast pulses,offers a new dimension for the regulation of quantum optical fields.In this paper,we established a temporal Fourier transform(TF)system based on ...Temporal optics,which enables lossless manipulation of ultrafast pulses,offers a new dimension for the regulation of quantum optical fields.In this paper,we established a temporal Fourier transform(TF)system based on a four-wave mixing(FWM)time lens and constructed a full quantum theoretical model for the resulting temporal SU(1,1)interferometer.This interferometer has high temporal resolution,can impose interference in both time and frequency domains,and is sensitive to the phase derivative.By introducing linear time-varying phase modulation,we achieved sub-picosecond precision in temporal autocorrelation measurements and generatedan optical frequency comb with a fixed interval based on a feedback iteration mechanism.Theoretical analysis revealsthe crucial regulatory role of time-frequency coupling in quantum interference,providing novel solutions for ultrafast quantum imaging,temporal mode encoding,and the generation of optical frequency quantization.展开更多
Background Lip reading uses lip images for visual speech recognition.Deep-learning-based lip reading has greatly improved performance in current datasets;however,most existing research ignores the significance of shor...Background Lip reading uses lip images for visual speech recognition.Deep-learning-based lip reading has greatly improved performance in current datasets;however,most existing research ignores the significance of short-term temporal dependencies of lip-shape variations between adjacent frames,which leaves space for further improvement in feature extraction.Methods This article presents a spatiotemporal feature fusion network(STDNet)that compensates for the deficiencies of current lip-reading approaches in short-term temporal dependency modeling.Specifically,to distinguish more similar and intricate content,STDNet adds a temporal feature extraction branch based on a 3D-CNN,which enhances the learning of dynamic lip movements in adjacent frames while not affecting spatial feature extraction.In particular,we designed a local–temporal block,which aggregates interframe differences,strengthening the relationship between various local lip regions through multiscale convolution.We incorporated the squeeze-and-excitation mechanism into the Global-Temporal Block,which processes a single frame as an independent unitto learn temporal variations across the entire lip region more effectively.Furthermore,attention pooling was introduced to highlight meaningful frames containing key semantic information for the target word.Results Experimental results demonstrated STDNet's superior performance on the LRW and LRW-1000,achieving word-level recognition accuracies of 90.2% and 53.56%,respectively.Extensive ablation experiments verified the rationality and effectiveness of its modules.Conclusions The proposed model effectively addresses short-term temporal dependency limitations in lip reading,and improves the temporal robustness of the model against variable-length sequences.These advancements validate the importance of explicit short-term dynamics modeling for practical lip-reading systems.展开更多
The Yellow River Basin in Sichuan Province(YRS)is undergoing severe soil erosion and exacerbated ecological vulnerability,which collectively pose formidable challenges for regional water conservation(WC)and sustainabl...The Yellow River Basin in Sichuan Province(YRS)is undergoing severe soil erosion and exacerbated ecological vulnerability,which collectively pose formidable challenges for regional water conservation(WC)and sustainable development.While effectively enhancing WC necessitates a comprehensive understanding of its driving factors and corresponding intervention strategies,existing studies have largely neglected the spatiotemporal heterogeneity of both natural and socio-economic drivers.Therefore,this study explored the spatiotemporal heterogeneity of WC drivers in YRS using multi-scale geographically weighted regression(MGWR)and geographically and temporally weighted regression(GTWR)models from an eco-hydrological perspective.We discovered that downstream regions,which are more developed,achieved significantly better WC than upstream regions.The results also demonstrated that the influence of temperature and wind speed is consistently dominant and temporally stable due to climate stability,while the influence of vegetation shifted from negative to positive around 2010,likely indicating greater benefits from understory vegetation.Economic growth positively impacted WC in upstream regions but had a negative effect in the more developed downstream regions.These findings highlight the importance of targeted water conservation strategies,including locally appropriate revegetation,optimization of agricultural and economic structures,and the establishment of eco-compensation mechanisms for ecological conservation and sustainable development.展开更多
Background Cervical cancer is the only cancer that can be eliminated worldwide.Tracking the latest burden of cervical cancer is critical toward the targets set by World Health Organization(WHO)to eliminate cervical ca...Background Cervical cancer is the only cancer that can be eliminated worldwide.Tracking the latest burden of cervical cancer is critical toward the targets set by World Health Organization(WHO)to eliminate cervical cancer as a major public health problem.Methods All data were extracted from the Global Cancer Observatory(GLOBOCAN)2022.Age-standardized incidence rate(ASIR)and mortality rates(ASMR)of cervical cancer were compared and linked to Human Development Index(HDI)between populations.The estimated annual percentage changes(EAPCs)were used to characterize the temporal trend in ASIR/ASMR,and demographic estimates were projected up to 2050.Results Globally,an estimated 662,044 cases(ASIR:14.12/100,000)and 348,709 deaths(ASMR:7.08/100,000)from cervical cancer occurred in 2022,corresponding to the fourth cause of cancer morbidity and mortality in women worldwide.Specifically,42%of cases and 39%of deaths occurred in China(23%and 16%)and India(19%and 23%).Both ASIR and ASMR of cervical cancer decreased with HDI,and similar decreasing links were observed for both early-onset(0–39 years)and late-onset(≥40 years)cervical cancer.Both ASIR and ASMR of overall cervical cancer showed decreasing trends during 2003–2012(EAPC:0.04%and-1.03%);however,upward trends were observed for early-onset cervical cancer(EAPC:1.16%and 0.57%).If national rates in 2022 remain stable,the estimated cases and deaths from cervical cancer are projected to increase by 56.8%and 80.7%up to 2050.Moreover,the projected increase of early-onset cervical cancer is mainly observed in transitioning countries,while decreased burden is expected in transitioned countries.Conclusions Cervical cancer remains a common cause of cancer death in many countries,especially in transitioning countries.Unless scaling-up preventive interventions,human papillomavirus(HPV)vaccination and cervical cancer screening,as well as systematic cooperation within government,civil societies,and private enterprises,the global burden of cervical cancer would be expected to increase in the future.展开更多
Mineralogical data are presented for the peridotite xenoliths from Miocene(~19 Ma)Qingyuan basalts in the eastern North China Craton(NCC),with the aim of constraining on property of the sub-continental lithospheric ma...Mineralogical data are presented for the peridotite xenoliths from Miocene(~19 Ma)Qingyuan basalts in the eastern North China Craton(NCC),with the aim of constraining on property of the sub-continental lithospheric mantle(SCLM)beneath the northern Tan-Lu fault zone(TLFZ)during the Cenozoic.The Qingyuan peridotites are dominated by spinel lherzolites with moderate-Mg^(#)olivines(89.4 to 91.2),suggesting that the regional SCLM is mainly transitional and fertile.Light rare earth element(LREE)-depleted,slightly depleted and enriched clinopyroxenes(Cpx)are identified in different peridotites.Chemical compositions of the LREE-enriched Cpx and the presence of phlogopite suggest that the Qingyuan SCLM has experienced silicate-related metasomatism.The synthesis of available mineral chemical data of the mantle xenoliths across the NCC confirms the SCLM beneath the NCC is highly heterogeneous in time and space.The Mesozoic–Cenozoic SCLM beneath the TLFZ and neighboring regions are more fertile and thinner than that beneath the region away from the fault zone.The fertile and refractory peridotite xenoliths experienced varying degrees of silicate and carbonatite metasomatism,respectively.The spatial-temporal lithospheric mantle heterogeneity in composition,age and thickness suggest that the trans-lithosphere fault zone played an important role in heterogeneous replacement of refractory cratonic lithospheric mantle.展开更多
Video snapshot compressive imaging(Video SCI) modulates scenes using various encoding masks and captures compressed measurements with a low-speed camera during a single exposure. Subsequently, reconstruction algorithm...Video snapshot compressive imaging(Video SCI) modulates scenes using various encoding masks and captures compressed measurements with a low-speed camera during a single exposure. Subsequently, reconstruction algorithms restore image sequences of dynamic scenes, offering advantages such as reduced bandwidth and storage space requirements. The temporal correlation in video data is crucial for Video SCI, as it leverages the temporal relationships among frames to enhance the efficiency and quality of reconstruction algorithms, particularly for fast-moving objects.This paper discretizes video frames to create image datasets with the same data volume but differing temporal correlations. We utilized the state-of-the-art(SOTA) reconstruction framework, EfficientSCI++, to train various compressed reconstruction models with these differing temporal correlations. Evaluating the reconstruction results from these models, our simulation experiments confirm that a reduction in temporal correlation leads to decreased reconstruction accuracy. Additionally, we simulated the reconstruction outcomes of datasets devoid of temporal correlation, illustrating that models trained on non-temporal data affect the temporal feature extraction capabilities of transformers, resulting in negligible impacts on the evaluation of reconstruction results for non-temporal correlation test datasets.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
To enhance the denoising performance of event-based sensors,we introduce a clustering-based temporal deep neural network denoising method(CBTDNN).Firstly,to cluster the sensor output data and obtain the respective clu...To enhance the denoising performance of event-based sensors,we introduce a clustering-based temporal deep neural network denoising method(CBTDNN).Firstly,to cluster the sensor output data and obtain the respective cluster centers,a combination of density-based spatial clustering of applications with noise(DBSCAN)and Kmeans++is utilized.Subsequently,long short-term memory(LSTM)is employed to fit and yield optimized cluster centers with temporal information.Lastly,based on the new cluster centers and denoising ratio,a radius threshold is set,and noise points beyond this threshold are removed.The comprehensive denoising metrics F1_score of CBTDNN have achieved 0.8931,0.7735,and 0.9215 on the traffic sequences dataset,pedestrian detection dataset,and turntable dataset,respectively.And these metrics demonstrate improvements of 49.90%,33.07%,19.31%,and 22.97%compared to four contrastive algorithms,namely nearest neighbor(NNb),nearest neighbor with polarity(NNp),Autoencoder,and multilayer perceptron denoising filter(MLPF).These results demonstrate that the proposed method enhances the denoising performance of event-based sensors.展开更多
Background Given the relatively unfavorable prognosis and significant geographic differences in lung cancer burden,it is critical to update the global landscape of lung cancer to inform local strategies.Methods Based ...Background Given the relatively unfavorable prognosis and significant geographic differences in lung cancer burden,it is critical to update the global landscape of lung cancer to inform local strategies.Methods Based on the GLOBOCAN 2022,the age-standardized incidence rate(ASIR)and mortality rate(ASMR)were compared and linked to the Human Development Index(HDI)across different populations.The temporal trends in ASIR/ASMR were characterized as estimated annual percentage change(EAPC),and demographic projections were performed up to 2050.Results Globally,an estimated 2,480,675 cases and 1,817,469 deaths from lung cancer occurred in 2022.Both ASIR and ASMR of lung cancer varied widely by world region,with ASIR ranging from 2.06 to 39.38 per 100,000 and ASMR from 1.95 to 31.70 per 100,000.China alone accounted for>40%of cases and deaths worldwide.Both ASIR and ARMR of lung cancer increased with HDI(R2:0.54 and 0.47,all P values<0.001),regardless of gender.Based on available data,both ASIR during 2001–2010 and ASMR during 2001–2015 showed decreasing trends in males(EAPC:1.50%and−2.22%)but increasing trends in females(EAPC:1.08%and 0.07%).Similar trends in ASIR and ASMR were observed among the elder population(≥50 years);however,downward trends were observed in the younger population(<50 years).Alongside the aging and growth of the population,estimated cases and deaths from overall lung cancer would increase by 86.2%and 95.2%up to 2050 as compared with estimates in 2022,respectively.Notably,increased early-onset lung cancer was only observed in transitioning countries,while decreased early-onset lung cancer was observed in transitioned countries.Conclusion Lung cancer maintained as the leading cancer burden worldwide.Unless timely preventive interventions in tobacco mitigation,early screening,and precise treatment,the global lung cancer burden is expected to increase in the future,especially for transitioning countries.展开更多
BACKGROUND Gastric bezoars are masses of indigestible material that accumulate in the stomach,causing nausea,abdominal pain,and vomiting.Persimmon bezoars(diospyrobezoars),which comprise tannins and fibers from persim...BACKGROUND Gastric bezoars are masses of indigestible material that accumulate in the stomach,causing nausea,abdominal pain,and vomiting.Persimmon bezoars(diospyrobezoars),which comprise tannins and fibers from persimmons,are relatively rare but may cause significant gastric complications,including gastric outlet obstruction or ileus.Although computed tomography(CT)is a useful ima-ging tool,diagnosing bezoars can be challenging because their density is similar to that of food debris and gastric content.CASE SUMMARY Here,we report the case of a 72-year-old woman with a persimmon bezoar that was diagnosed using serial CT imaging and confirmed by endoscopy.CT perfor-med over several months revealed changes in the internal structure and density of the bezoar,suggesting progressive hardening.The patient had a history of a par-tial gastrectomy and excessive persimmon consumption,both of which are risk factors for bezoar formation.Endoscopic fragmentation of the bezoar successfully resolved symptoms.CONCLUSION Gastric bezoars,particularly persimmon bezoars,present diagnostic challenges because of their variable imaging characteristics.Serial CT can document tem-poral changes in bezoar density,potentially reflecting changes in hardness.Early diagnosis and endoscopic treatment are essential for effective management,particularly in patients with predisposing factors.This case underscores the im-portance of considering bezoars in the differential diagnosis of gastric masses,and highlights the value of CT for monitoring changes in bezoar characteristics over time.展开更多
High-resolution spatiotemporal simulations effectively capture the complexities of atmospheric plume sion disper-in complex terrain.However,their high computational cost makes them impractical for applications requiri...High-resolution spatiotemporal simulations effectively capture the complexities of atmospheric plume sion disper-in complex terrain.However,their high computational cost makes them impractical for applications requiring rapid responses or iterative processes,such as optimization,uncertainty quantification,or inverse modeling.To address this challenge,this work introduces the Dual-Stage Temporal Three-dimensional UNet Super-resolution(DST3D-UNet-SR)model,a highly efficient deep learning model for plume dispersion predictions.DST3D-UNet-SR is composed of two sequential modules:the temporal module(TM),which predicts the transient evolution of a plume in complex terrain from low-resolution temporal data,and the spatial refinement module(SRM),which subsequently enhances the spatial resolution of the TM predictions.We train DST3D-UNet-SR using a comprehensive dataset derived from high-resolution large eddy simulations(LES)of plume transport.We propose the DST3D-UNet-SR model to significantly accelerate LES of three-dimensional(3D)plume dispersion by three orders of magnitude.Additionally,the model demonstrates the ability to dynamically adapt to evolving conditions through the incorporation of new observational data,substantially improving prediction accuracy in high-concentration regions near the source.展开更多
Tuberculosis(TB)remained the first leading cause of death from a single infectious agent worldwide in 2023,resulting in nearly twice as many deaths as those caused by the human immunodeficiency virus/acquired immune d...Tuberculosis(TB)remained the first leading cause of death from a single infectious agent worldwide in 2023,resulting in nearly twice as many deaths as those caused by the human immunodeficiency virus/acquired immune deficiency syndrome.An estimated 10.8 million TB cases were reported globally in 2023,with approximately 1.25 million associated deaths.In China,which ranks third in the global TB burden,there were approximately 741,000 new cases and 25,000 deaths in 2023^([1]).TB poses a significant threat to human health worldwide.展开更多
Approximately 30%-40% of epilepsy patients do not respond well to adequate anti-seizure medications(ASMs),a condition known as pharmacoresistant epilepsy.The management of pharmacoresistant epilepsy remains an intract...Approximately 30%-40% of epilepsy patients do not respond well to adequate anti-seizure medications(ASMs),a condition known as pharmacoresistant epilepsy.The management of pharmacoresistant epilepsy remains an intractable issue in the clinic.Its early prediction is important for prevention and diagnosis.However,it still lacks effective predictors and approaches.Here,a classical model of pharmacoresistant temporal lobe epilepsy(TLE)was established to screen pharmacoresistant and pharmaco-responsive individuals by applying phenytoin to amygdaloid-kindled rats.Ictal electroencephalograms(EEGs)recorded before phenytoin treatment were analyzed.Based on ictal EEGs from pharmacoresistant and pharmaco-responsive rats,a convolutional neural network predictive model was constructed to predict pharmacoresistance,and achieved 78% prediction accuracy.We further found the ictal EEGs from pharmacoresistant rats have a lower gamma-band power,which was verified in seizure EEGs from pharmacoresistant TLE patients.Prospectively,therapies targeting the subiculum in those predicted as“pharmacoresistant”individual rats significantly reduced the subsequent occurrence of pharmacoresistance.These results demonstrate a new methodology to predict whether TLE individuals become resistant to ASMs in a classic pharmacoresistant TLE model.This may be of translational importance for the precise management of pharmacoresistant TLE.展开更多
Recently,information acquired at the canopy top,such as spectral and textural data,has been widely used to estimate plant nitrogen(N)accumulation(PNA).The response of crops to N uptake involves not only changes in hor...Recently,information acquired at the canopy top,such as spectral and textural data,has been widely used to estimate plant nitrogen(N)accumulation(PNA).The response of crops to N uptake involves not only changes in horizontal canopy top information but also an increase in vertical plant height(PH).It remains unclear whether the fusion of spectral indices with PH can improve the estimation performance of PNA models based on spectral remote sensing across different growth stages.展开更多
Based on the monitoring data of PM 10 concentration from six environmental monitoring stations and the ground meteorological observation data in Yantai City from 2019 to 2021,the spatial and temporal variation of PM 1...Based on the monitoring data of PM 10 concentration from six environmental monitoring stations and the ground meteorological observation data in Yantai City from 2019 to 2021,the spatial and temporal variation of PM 10 concentration and its relationship with meteorological factors were studied.The results show that from the perspective of temporal variation,the annual average of PM 10 concentration in Yantai City tended to decrease year by year.It was high in winter and spring and low in summer and autumn.In terms of monthly variation,the changing curve is U-shaped,and it was high in December and January but low in July and August.During a day,PM 10 concentration had two peaks.The first peak appeared approximately from 09:00 to 11:00,and the second peak can be found from 21:00 to 23:00.From the perspective of spatial distribution,PM 10 concentration was the highest in the development area and Fushan District.It was the highest in the west,followed by the east,while it was the lowest in the middle.The spatial difference rate was the highest in summer.Average temperature,relative humidity,wind speed and precipitation were the main meteorological factors influencing PM 10 concentration in Yantai area.PM 10 concentration was negatively correlated with average temperature and relative humidity,and the correlation was the most significant from June to October.It was negatively correlated with wind speed and precipitation,and the correlation was different in various months.The negative correlation was significant in summer and winter.展开更多
基金supported by the National Key R&D Program of China(No.2018YFB1305200)the Natural Science Foundation of Zhejiang Province(No.LGG21F030011)。
文摘Research on human motion prediction has made significant progress due to its importance in the development of various artificial intelligence applications.However,effectively capturing spatio-temporal features for smoother and more precise human motion prediction remains a challenge.To address these issues,a robust human motion prediction method via integration of spatial and temporal cues(RISTC)has been proposed.This method captures sufficient spatio-temporal correlation of the observable sequence of human poses by utilizing the spatio-temporal mixed feature extractor(MFE).In multi-layer MFEs,the channel-graph united attention blocks extract the augmented spatial features of the human poses in the channel and spatial dimension.Additionally,multi-scale temporal blocks have been designed to effectively capture complicated and highly dynamic temporal information.Our experiments on the Human3.6M and Carnegie Mellon University motion capture(CMU Mocap)datasets show that the proposed network yields higher prediction accuracy than the state-of-the-art methods.
基金The National Key Research and Development Program of China(No.2022YFC3800201).
文摘This study examined the influence of the built environment surrounding rail stations on rail transit ridership and its spatiotemporal variations,aiming to enhance rail transit operational efficiency and inform station planning and development.Data from 159 metro stations in Nanjing,collected over a 14-d period,were analyzed to identify changes in weekday and weekend ridership patterns.The analysis included explanatory variables grouped into three categories:urban spatial variables,socioeconomic vari-ables,and transit service variables.A geographically and temporally weighted regression(GTWR)model was developed,and its performance was compared with that of ordinary least squares(OLS)and geographically weighted regression(GWR)models.The results demonstrated that the GTWR model outperformed others in analyzing the relationship between rail transit ridership and the built environment.In addition,the coefficients of explanatory variables showed significant variation across spatiotemporal dimensions,revealing distinct patterns.Notably,the influence of commuter flows led to more pronounced temporal heterogeneity in the coefficients observed on weekdays.These findings offer valuable insights for optimizing urban public transportation systems and advancing integrated urban rail development.
基金Shanghai Municipal Commission of Economy and Information Technology,China (No.202301054)。
文摘Action recognition,a fundamental task in the field of video understanding,has been extensively researched and applied.In contrast to an image,a video introduces an extra temporal dimension.However,many existing action recognition networks either perform simple temporal fusion through averaging or rely on pre-trained models from image recognition,resulting in limited temporal information extraction capabilities.This work proposes a highly efficient temporal decoding module that can be seamlessly integrated into any action recognition backbone network to enhance the focus on temporal relationships between video frames.Firstly,the decoder initializes a set of learnable queries,termed video-level action category prediction queries.Then,they are combined with the video frame features extracted by the backbone network after self-attention learning to extract video context information.Finally,these prediction queries with rich temporal features are used for category prediction.Experimental results on HMDB51,MSRDailyAct3D,Diving48 and Breakfast datasets show that using TokShift-Transformer and VideoMAE as encoders results in a significant improvement in Top-1 accuracy compared to the original models(TokShift-Transformer and VideoMAE),after introducing the proposed temporal decoder.The introduction of the temporal decoder results in an average performance increase exceeding 11%for TokShift-Transformer and nearly 5%for VideoMAE across the four datasets.Furthermore,the work explores the combination of the decoder with various action recognition networks,including Timesformer,as encoders.This results in an average accuracy improvement of more than 3.5%on the HMDB51 dataset.The code is available at https://github.com/huangturbo/TempDecoder.
文摘The online version of the original article can be found at:https://www.sciopen.com/article/10.26599/JOTO.2025.9540018 Erratum to Journal of Otology,2025,20(2):123-126.https://doi.org/10.26599/JOTO.2025.9540018 The surnames and given names of these authors are reversed:Saro-Buendía Miguel,Andresen-Lorca Belén,Pérez-García Alberto,Llópez Carratala Nacho,Carreres Polo Joan,Armengot Carceller Miguel,Perolada Valmaña Jose María.It should be Miguel Saro-Buendía,Belén Andresen-Lorca,Alberto Pérez-García,Nacho Llópez Carratala,Joan Carreres Polo,Miguel Armengot Carceller,Jose María Perolada Valmaña.
文摘Osteoradionecrosis of the temporal bone(ORN-TB)is usually controlled with conservative measures.However,a temporal bone resection may be required in unresponsive cases.The reconstruction of the resulting defects may be challenging because of the radiation damage to regional tissues.As a result,distant free flaps may be an optimal choice.For instance,the gracilis muscular free flap(GMFF)has consistent vascular anatomy and can be used to reconstruct small defects.We report three cases of uncontrolled ORN-TB requiring an extensive temporal bone resection followed by vascularized obliteration with a GMFF.The patients reported complete control of the main otologic symptoms(otorrhea,otalgia,and aural fullness)and optimal functional and aesthetic outcomes.Finally,the patients reported significant improvement in quality of life despite early postoperative complications.To our knowledge,the GMFF had not been used to obliterate temporal bone defects in patients with ORN-TB.
文摘Temporal optics,which enables lossless manipulation of ultrafast pulses,offers a new dimension for the regulation of quantum optical fields.In this paper,we established a temporal Fourier transform(TF)system based on a four-wave mixing(FWM)time lens and constructed a full quantum theoretical model for the resulting temporal SU(1,1)interferometer.This interferometer has high temporal resolution,can impose interference in both time and frequency domains,and is sensitive to the phase derivative.By introducing linear time-varying phase modulation,we achieved sub-picosecond precision in temporal autocorrelation measurements and generatedan optical frequency comb with a fixed interval based on a feedback iteration mechanism.Theoretical analysis revealsthe crucial regulatory role of time-frequency coupling in quantum interference,providing novel solutions for ultrafast quantum imaging,temporal mode encoding,and the generation of optical frequency quantization.
基金Supported by the National Key Research and Development Program of China(2023YFC3306201)the National Natural Science Foundation of China(61772125)the Fundamental Research Funds for the Central Universities(N2317004).
文摘Background Lip reading uses lip images for visual speech recognition.Deep-learning-based lip reading has greatly improved performance in current datasets;however,most existing research ignores the significance of short-term temporal dependencies of lip-shape variations between adjacent frames,which leaves space for further improvement in feature extraction.Methods This article presents a spatiotemporal feature fusion network(STDNet)that compensates for the deficiencies of current lip-reading approaches in short-term temporal dependency modeling.Specifically,to distinguish more similar and intricate content,STDNet adds a temporal feature extraction branch based on a 3D-CNN,which enhances the learning of dynamic lip movements in adjacent frames while not affecting spatial feature extraction.In particular,we designed a local–temporal block,which aggregates interframe differences,strengthening the relationship between various local lip regions through multiscale convolution.We incorporated the squeeze-and-excitation mechanism into the Global-Temporal Block,which processes a single frame as an independent unitto learn temporal variations across the entire lip region more effectively.Furthermore,attention pooling was introduced to highlight meaningful frames containing key semantic information for the target word.Results Experimental results demonstrated STDNet's superior performance on the LRW and LRW-1000,achieving word-level recognition accuracies of 90.2% and 53.56%,respectively.Extensive ablation experiments verified the rationality and effectiveness of its modules.Conclusions The proposed model effectively addresses short-term temporal dependency limitations in lip reading,and improves the temporal robustness of the model against variable-length sequences.These advancements validate the importance of explicit short-term dynamics modeling for practical lip-reading systems.
基金supported by the funding provided by the State Key Laboratory of Hydraulics and Mountain River Engineering(SKHL2210)National Natural Science Foundation of China(42171304)+1 种基金the Sichuan Science and Technology Program(2023YFS0380)Natural Science Foundation of Jiangsu Province of China(BK20242018)。
文摘The Yellow River Basin in Sichuan Province(YRS)is undergoing severe soil erosion and exacerbated ecological vulnerability,which collectively pose formidable challenges for regional water conservation(WC)and sustainable development.While effectively enhancing WC necessitates a comprehensive understanding of its driving factors and corresponding intervention strategies,existing studies have largely neglected the spatiotemporal heterogeneity of both natural and socio-economic drivers.Therefore,this study explored the spatiotemporal heterogeneity of WC drivers in YRS using multi-scale geographically weighted regression(MGWR)and geographically and temporally weighted regression(GTWR)models from an eco-hydrological perspective.We discovered that downstream regions,which are more developed,achieved significantly better WC than upstream regions.The results also demonstrated that the influence of temperature and wind speed is consistently dominant and temporally stable due to climate stability,while the influence of vegetation shifted from negative to positive around 2010,likely indicating greater benefits from understory vegetation.Economic growth positively impacted WC in upstream regions but had a negative effect in the more developed downstream regions.These findings highlight the importance of targeted water conservation strategies,including locally appropriate revegetation,optimization of agricultural and economic structures,and the establishment of eco-compensation mechanisms for ecological conservation and sustainable development.
基金supported by the National Key R&D Program of China(grant number:2021YFC2500400)National Natural Science Foundation of China(grant numbers:82172894,82073028,82204121)China Postdoctoral Science Foundation(grant number:2023M742617).
文摘Background Cervical cancer is the only cancer that can be eliminated worldwide.Tracking the latest burden of cervical cancer is critical toward the targets set by World Health Organization(WHO)to eliminate cervical cancer as a major public health problem.Methods All data were extracted from the Global Cancer Observatory(GLOBOCAN)2022.Age-standardized incidence rate(ASIR)and mortality rates(ASMR)of cervical cancer were compared and linked to Human Development Index(HDI)between populations.The estimated annual percentage changes(EAPCs)were used to characterize the temporal trend in ASIR/ASMR,and demographic estimates were projected up to 2050.Results Globally,an estimated 662,044 cases(ASIR:14.12/100,000)and 348,709 deaths(ASMR:7.08/100,000)from cervical cancer occurred in 2022,corresponding to the fourth cause of cancer morbidity and mortality in women worldwide.Specifically,42%of cases and 39%of deaths occurred in China(23%and 16%)and India(19%and 23%).Both ASIR and ASMR of cervical cancer decreased with HDI,and similar decreasing links were observed for both early-onset(0–39 years)and late-onset(≥40 years)cervical cancer.Both ASIR and ASMR of overall cervical cancer showed decreasing trends during 2003–2012(EAPC:0.04%and-1.03%);however,upward trends were observed for early-onset cervical cancer(EAPC:1.16%and 0.57%).If national rates in 2022 remain stable,the estimated cases and deaths from cervical cancer are projected to increase by 56.8%and 80.7%up to 2050.Moreover,the projected increase of early-onset cervical cancer is mainly observed in transitioning countries,while decreased burden is expected in transitioned countries.Conclusions Cervical cancer remains a common cause of cancer death in many countries,especially in transitioning countries.Unless scaling-up preventive interventions,human papillomavirus(HPV)vaccination and cervical cancer screening,as well as systematic cooperation within government,civil societies,and private enterprises,the global burden of cervical cancer would be expected to increase in the future.
基金supported by funds from the Ministry of Science and Technology of the People's Republic of China(No.2019YFA0708603)NSFC(Nos.41973050,42288201,41930215)the Key Special Project for Introduced Talents Team of Southern Marine Science and Engineering Guangdong Laboratory(Guangzhou)(No.GML2019ZD0202)。
文摘Mineralogical data are presented for the peridotite xenoliths from Miocene(~19 Ma)Qingyuan basalts in the eastern North China Craton(NCC),with the aim of constraining on property of the sub-continental lithospheric mantle(SCLM)beneath the northern Tan-Lu fault zone(TLFZ)during the Cenozoic.The Qingyuan peridotites are dominated by spinel lherzolites with moderate-Mg^(#)olivines(89.4 to 91.2),suggesting that the regional SCLM is mainly transitional and fertile.Light rare earth element(LREE)-depleted,slightly depleted and enriched clinopyroxenes(Cpx)are identified in different peridotites.Chemical compositions of the LREE-enriched Cpx and the presence of phlogopite suggest that the Qingyuan SCLM has experienced silicate-related metasomatism.The synthesis of available mineral chemical data of the mantle xenoliths across the NCC confirms the SCLM beneath the NCC is highly heterogeneous in time and space.The Mesozoic–Cenozoic SCLM beneath the TLFZ and neighboring regions are more fertile and thinner than that beneath the region away from the fault zone.The fertile and refractory peridotite xenoliths experienced varying degrees of silicate and carbonatite metasomatism,respectively.The spatial-temporal lithospheric mantle heterogeneity in composition,age and thickness suggest that the trans-lithosphere fault zone played an important role in heterogeneous replacement of refractory cratonic lithospheric mantle.
基金supported in part by the National Natural Science Foundation of China (No. U23B2011)。
文摘Video snapshot compressive imaging(Video SCI) modulates scenes using various encoding masks and captures compressed measurements with a low-speed camera during a single exposure. Subsequently, reconstruction algorithms restore image sequences of dynamic scenes, offering advantages such as reduced bandwidth and storage space requirements. The temporal correlation in video data is crucial for Video SCI, as it leverages the temporal relationships among frames to enhance the efficiency and quality of reconstruction algorithms, particularly for fast-moving objects.This paper discretizes video frames to create image datasets with the same data volume but differing temporal correlations. We utilized the state-of-the-art(SOTA) reconstruction framework, EfficientSCI++, to train various compressed reconstruction models with these differing temporal correlations. Evaluating the reconstruction results from these models, our simulation experiments confirm that a reduction in temporal correlation leads to decreased reconstruction accuracy. Additionally, we simulated the reconstruction outcomes of datasets devoid of temporal correlation, illustrating that models trained on non-temporal data affect the temporal feature extraction capabilities of transformers, resulting in negligible impacts on the evaluation of reconstruction results for non-temporal correlation test datasets.
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supported by the National Natural Science Foundation of China(No.62134004).
文摘To enhance the denoising performance of event-based sensors,we introduce a clustering-based temporal deep neural network denoising method(CBTDNN).Firstly,to cluster the sensor output data and obtain the respective cluster centers,a combination of density-based spatial clustering of applications with noise(DBSCAN)and Kmeans++is utilized.Subsequently,long short-term memory(LSTM)is employed to fit and yield optimized cluster centers with temporal information.Lastly,based on the new cluster centers and denoising ratio,a radius threshold is set,and noise points beyond this threshold are removed.The comprehensive denoising metrics F1_score of CBTDNN have achieved 0.8931,0.7735,and 0.9215 on the traffic sequences dataset,pedestrian detection dataset,and turntable dataset,respectively.And these metrics demonstrate improvements of 49.90%,33.07%,19.31%,and 22.97%compared to four contrastive algorithms,namely nearest neighbor(NNb),nearest neighbor with polarity(NNp),Autoencoder,and multilayer perceptron denoising filter(MLPF).These results demonstrate that the proposed method enhances the denoising performance of event-based sensors.
基金supported by the National Key Research and Development Program of China(grant number:2021YFC2500400)Tianjin Health Committee Foundation(grant number:TJWJ2021MS008)+2 种基金Tianjin Key Medical Discipline(Specialty)Construction Project(grant number:TYXZDXK-009A)Science and Technology Program of the Joint Fund of Scientific Research for the Public Hospitals of Inner Mongolia Academy of Medical Sciences(grant number:2023GLLH0132)Scientific Research Fund for the Demonstration Project of Public Hospital Reform and Quality Development(Gastrointestinal Tumour)that is approved by Peking University Cancer Hospital(Inner Mongolia Campus)(grant number:2023SGGZ068)。
文摘Background Given the relatively unfavorable prognosis and significant geographic differences in lung cancer burden,it is critical to update the global landscape of lung cancer to inform local strategies.Methods Based on the GLOBOCAN 2022,the age-standardized incidence rate(ASIR)and mortality rate(ASMR)were compared and linked to the Human Development Index(HDI)across different populations.The temporal trends in ASIR/ASMR were characterized as estimated annual percentage change(EAPC),and demographic projections were performed up to 2050.Results Globally,an estimated 2,480,675 cases and 1,817,469 deaths from lung cancer occurred in 2022.Both ASIR and ASMR of lung cancer varied widely by world region,with ASIR ranging from 2.06 to 39.38 per 100,000 and ASMR from 1.95 to 31.70 per 100,000.China alone accounted for>40%of cases and deaths worldwide.Both ASIR and ARMR of lung cancer increased with HDI(R2:0.54 and 0.47,all P values<0.001),regardless of gender.Based on available data,both ASIR during 2001–2010 and ASMR during 2001–2015 showed decreasing trends in males(EAPC:1.50%and−2.22%)but increasing trends in females(EAPC:1.08%and 0.07%).Similar trends in ASIR and ASMR were observed among the elder population(≥50 years);however,downward trends were observed in the younger population(<50 years).Alongside the aging and growth of the population,estimated cases and deaths from overall lung cancer would increase by 86.2%and 95.2%up to 2050 as compared with estimates in 2022,respectively.Notably,increased early-onset lung cancer was only observed in transitioning countries,while decreased early-onset lung cancer was observed in transitioned countries.Conclusion Lung cancer maintained as the leading cancer burden worldwide.Unless timely preventive interventions in tobacco mitigation,early screening,and precise treatment,the global lung cancer burden is expected to increase in the future,especially for transitioning countries.
文摘BACKGROUND Gastric bezoars are masses of indigestible material that accumulate in the stomach,causing nausea,abdominal pain,and vomiting.Persimmon bezoars(diospyrobezoars),which comprise tannins and fibers from persimmons,are relatively rare but may cause significant gastric complications,including gastric outlet obstruction or ileus.Although computed tomography(CT)is a useful ima-ging tool,diagnosing bezoars can be challenging because their density is similar to that of food debris and gastric content.CASE SUMMARY Here,we report the case of a 72-year-old woman with a persimmon bezoar that was diagnosed using serial CT imaging and confirmed by endoscopy.CT perfor-med over several months revealed changes in the internal structure and density of the bezoar,suggesting progressive hardening.The patient had a history of a par-tial gastrectomy and excessive persimmon consumption,both of which are risk factors for bezoar formation.Endoscopic fragmentation of the bezoar successfully resolved symptoms.CONCLUSION Gastric bezoars,particularly persimmon bezoars,present diagnostic challenges because of their variable imaging characteristics.Serial CT can document tem-poral changes in bezoar density,potentially reflecting changes in hardness.Early diagnosis and endoscopic treatment are essential for effective management,particularly in patients with predisposing factors.This case underscores the im-portance of considering bezoars in the differential diagnosis of gastric masses,and highlights the value of CT for monitoring changes in bezoar characteristics over time.
文摘High-resolution spatiotemporal simulations effectively capture the complexities of atmospheric plume sion disper-in complex terrain.However,their high computational cost makes them impractical for applications requiring rapid responses or iterative processes,such as optimization,uncertainty quantification,or inverse modeling.To address this challenge,this work introduces the Dual-Stage Temporal Three-dimensional UNet Super-resolution(DST3D-UNet-SR)model,a highly efficient deep learning model for plume dispersion predictions.DST3D-UNet-SR is composed of two sequential modules:the temporal module(TM),which predicts the transient evolution of a plume in complex terrain from low-resolution temporal data,and the spatial refinement module(SRM),which subsequently enhances the spatial resolution of the TM predictions.We train DST3D-UNet-SR using a comprehensive dataset derived from high-resolution large eddy simulations(LES)of plume transport.We propose the DST3D-UNet-SR model to significantly accelerate LES of three-dimensional(3D)plume dispersion by three orders of magnitude.Additionally,the model demonstrates the ability to dynamically adapt to evolving conditions through the incorporation of new observational data,substantially improving prediction accuracy in high-concentration regions near the source.
文摘Tuberculosis(TB)remained the first leading cause of death from a single infectious agent worldwide in 2023,resulting in nearly twice as many deaths as those caused by the human immunodeficiency virus/acquired immune deficiency syndrome.An estimated 10.8 million TB cases were reported globally in 2023,with approximately 1.25 million associated deaths.In China,which ranks third in the global TB burden,there were approximately 741,000 new cases and 25,000 deaths in 2023^([1]).TB poses a significant threat to human health worldwide.
基金supported by grants from the National Key R&D Program of China(2020YFA0803900)the National Natural Science Foundation of China(82173796 and U21A20418)+1 种基金the Natural Science Foundation of Zhejiang Province(LD22H310003)the Key R&D Plan of Zhejiang Province(2021C03116).
文摘Approximately 30%-40% of epilepsy patients do not respond well to adequate anti-seizure medications(ASMs),a condition known as pharmacoresistant epilepsy.The management of pharmacoresistant epilepsy remains an intractable issue in the clinic.Its early prediction is important for prevention and diagnosis.However,it still lacks effective predictors and approaches.Here,a classical model of pharmacoresistant temporal lobe epilepsy(TLE)was established to screen pharmacoresistant and pharmaco-responsive individuals by applying phenytoin to amygdaloid-kindled rats.Ictal electroencephalograms(EEGs)recorded before phenytoin treatment were analyzed.Based on ictal EEGs from pharmacoresistant and pharmaco-responsive rats,a convolutional neural network predictive model was constructed to predict pharmacoresistance,and achieved 78% prediction accuracy.We further found the ictal EEGs from pharmacoresistant rats have a lower gamma-band power,which was verified in seizure EEGs from pharmacoresistant TLE patients.Prospectively,therapies targeting the subiculum in those predicted as“pharmacoresistant”individual rats significantly reduced the subsequent occurrence of pharmacoresistance.These results demonstrate a new methodology to predict whether TLE individuals become resistant to ASMs in a classic pharmacoresistant TLE model.This may be of translational importance for the precise management of pharmacoresistant TLE.
基金supported by the National Key Research and Development Plan Project Sub-Topic of China(Grant Nos.2022YFD1901500 and 2022YFD1901505-07)the National Natural Science Foundation of China(Grant No.32260531)+1 种基金the Key Laboratory of Molecular Breeding for Grain and Oil Crops in Guizhou Province,China(Grant No.Qiankehezhongyindi[2023]8)the Key Laboratory of Functional Agriculture of Guizhou Provincial Higher Education Institutions,China(Grant No.Qianjiaoji[2023]007).
文摘Recently,information acquired at the canopy top,such as spectral and textural data,has been widely used to estimate plant nitrogen(N)accumulation(PNA).The response of crops to N uptake involves not only changes in horizontal canopy top information but also an increase in vertical plant height(PH).It remains unclear whether the fusion of spectral indices with PH can improve the estimation performance of PNA models based on spectral remote sensing across different growth stages.
基金the Science and Technology Research Project of Shandong Meteorological Bureau(2022SDQN11)Science and Technology Research Project of Yantai Meteorological Bureau(2024ytcx07).
文摘Based on the monitoring data of PM 10 concentration from six environmental monitoring stations and the ground meteorological observation data in Yantai City from 2019 to 2021,the spatial and temporal variation of PM 10 concentration and its relationship with meteorological factors were studied.The results show that from the perspective of temporal variation,the annual average of PM 10 concentration in Yantai City tended to decrease year by year.It was high in winter and spring and low in summer and autumn.In terms of monthly variation,the changing curve is U-shaped,and it was high in December and January but low in July and August.During a day,PM 10 concentration had two peaks.The first peak appeared approximately from 09:00 to 11:00,and the second peak can be found from 21:00 to 23:00.From the perspective of spatial distribution,PM 10 concentration was the highest in the development area and Fushan District.It was the highest in the west,followed by the east,while it was the lowest in the middle.The spatial difference rate was the highest in summer.Average temperature,relative humidity,wind speed and precipitation were the main meteorological factors influencing PM 10 concentration in Yantai area.PM 10 concentration was negatively correlated with average temperature and relative humidity,and the correlation was the most significant from June to October.It was negatively correlated with wind speed and precipitation,and the correlation was different in various months.The negative correlation was significant in summer and winter.
