Gaze estimation,a crucial non-verbal communication cue,has achieved remarkable progress through convolutional neural networks.However,accurate gaze prediction in uncon-strained environments,particularly in extreme hea...Gaze estimation,a crucial non-verbal communication cue,has achieved remarkable progress through convolutional neural networks.However,accurate gaze prediction in uncon-strained environments,particularly in extreme head poses,partial occlusions,and abnormal lighting,remains challenging.Existing models often struggle to effectively focus on discriminative ocular features,leading to suboptimal performance.To address these limitations,this paper proposes dual-branch gaze estimation with Gaussian mixture distribution heatmaps and dynamic adaptive loss function(DMGDL),a novel dual-branch gaze estimation algorithm.By introducing Gaussian mixture distribution heatmaps centered on pupil positions as spatial attention guides,the model is enabled to prioritize ocular regions.Additionally,a dual-branch network architecture is designed to separately extract features for yaw and pitch angles,enhancing flexibility and mitigating cross-angle interference.A dynamic adaptive loss function is further formulated to address discontinuities in angle estimation,improving robustness and convergence stability.Experimental evaluations on three benchmark datasets demonstrate that DMGDL outperforms state-of-the-art methods,achiev-ing a mean angular error of 3.98°on the Max-Planck institute for informatics face gaze(MPI-IFaceGaze)dataset,10.21°on the physically unconstrained gaze estimation in the wild(Gaze360)dataset and 6.14°on the real-time eye gaze estimation in natural environments(RT-Gene)dataset,exhibiting superior generalization and robustness.展开更多
Although compatibility is highly advocated in traditional Chinese medicine(TCM),inappropriate combination of some herbs may reduce the therapeutic action and even produce toxic effects.Kansui and licorice,one of TCM“...Although compatibility is highly advocated in traditional Chinese medicine(TCM),inappropriate combination of some herbs may reduce the therapeutic action and even produce toxic effects.Kansui and licorice,one of TCM“Eighteen Incompatible Medicaments”,are the most representative cases of improper herbal combination,which may still be applied simultaneously under given conditions.However,the potential mechanism of their compatibility and incompatibility is unclear.In the present study,two different ratios of kansui and licorice,representing their compatibility and incompatibility respectively,were designed to elucidate their interaction by comparative plasma/tissue metabolomics and a heatmap with relative fold change.As a result,glycocholic acid,prostaglandin F2a,dihydroceramide and sphinganine were screened out as the principal alternative biomarkers of compatibility group;sphinganine,dihydroceramide,arachidonic acid,leukotriene B4,acetoacetic acid and linoleic acid were those of incompatibility group.Based on the values of biomarkers in each tissue,the liver was identified as the compatible target organ,while the heart,liver,and kidney were the incompatible target organs.Furthermore,important pathways for compatibility and incompatibility were also constructed.These results help us to better understand and utilize the two herbs,and the study was the first to reveal some innate characters of herbs related to TCM“Eighteen Incompatible Medicaments”.展开更多
Recognition of human gesture actions is a challenging issue due to the complex patterns in both visual andskeletal features. Existing gesture action recognition (GAR) methods typically analyze visual and skeletal data...Recognition of human gesture actions is a challenging issue due to the complex patterns in both visual andskeletal features. Existing gesture action recognition (GAR) methods typically analyze visual and skeletal data,failing to meet the demands of various scenarios. Furthermore, multi-modal approaches lack the versatility toefficiently process both uniformand disparate input patterns.Thus, in this paper, an attention-enhanced pseudo-3Dresidual model is proposed to address the GAR problem, called HgaNets. This model comprises two independentcomponents designed formodeling visual RGB (red, green and blue) images and 3Dskeletal heatmaps, respectively.More specifically, each component consists of two main parts: 1) a multi-dimensional attention module forcapturing important spatial, temporal and feature information in human gestures;2) a spatiotemporal convolutionmodule that utilizes pseudo-3D residual convolution to characterize spatiotemporal features of gestures. Then,the output weights of the two components are fused to generate the recognition results. Finally, we conductedexperiments on four datasets to assess the efficiency of the proposed model. The results show that the accuracy onfour datasets reaches 85.40%, 91.91%, 94.70%, and 95.30%, respectively, as well as the inference time is 0.54 s andthe parameters is 2.74M. These findings highlight that the proposed model outperforms other existing approachesin terms of recognition accuracy.展开更多
Musculoskeletal injuries are among the most common causes of disability worldwide,with early detection and appropriate intervention critical to minimizing long-term complications.Infrared thermography(IRT)has emerged ...Musculoskeletal injuries are among the most common causes of disability worldwide,with early detection and appropriate intervention critical to minimizing long-term complications.Infrared thermography(IRT)has emerged as a noninvasive,real-time imaging modality that captures superficial temperature changes reflecting underlying physiological processes such as inflammation and vascular alterations.This review explores the fundamental principles of medical thermography,differentiates between passive and active approaches,and outlines key technological advancements including artificial intelligence integration.The clinical utility of IRT is discussed in various contexts–ranging from acute soft tissue injuries and overuse syndromes to chronic pain and rehabilitation monitoring.Comparative insights with conventional imaging techniques such as ultrasound and magnetic resonance imaging are also presented.While IRT offers functional imaging capabilities with advantages in portability,safety,and speed,its limitations–such as lack of deep-tissue penetration and protocol standardization–remain significant barriers to broader adoption.Future directions include the integration of IRT with other imaging modalities and digital health platforms to enhance musculoskeletal assessment and injury prevention strategies.展开更多
Flax(Linum usitatissimum L.)is an important oil crop in the high-altitude arid regions of China.Flaxseed is rich in various nutrients.However,the nutritional qualities of flaxseeds from different producing areas are s...Flax(Linum usitatissimum L.)is an important oil crop in the high-altitude arid regions of China.Flaxseed is rich in various nutrients.However,the nutritional qualities of flaxseeds from different producing areas are still unclear.In this study,the nutritional characteristics of flaxseed from five producing areas in China were investigated.Twenty five nutritional quality indices in flaxseed were analyzed.Subsequently,chemometric methods,including cluster analysis,principal component analysis(PCA)and partial least square discriminant analysis(PLS-DA),were employed to discover the characteristics of nutritional qualities in flaxseeds.The results revealed there are significant differences in nutritional qualities among flaxseeds from different production areas.Six quality indices includingγ-tocopherol,vitamin E,phytosterols,oleic acid,α-linolenic acid,and cycloartenol were susceptible to producing area.In detail,the superiorcharacteristic nutrients of Ningxia flaxseed,Inner Mongolia flaxseed and Hebei flaxseed are vitamin E(17.3 mg/100g),α-linolenic acid(52.6%)and cycloartenol(1738.1 mg/kg),and phytosterols(3032.0 mg/kg),respectively.This study promotes the high-value development and utilization of local flaxseed industry.展开更多
目的相比于一般视频目标检测跟踪任务,视频羽毛球的实时定位跟踪主要面临两大难点:1)羽毛球属于小目标,同时伴有严重的运动模糊以及相似目标的干扰,使用基于矩形框的目标检测跟踪方法准确率低且会带来中心点误差问题;2)单帧图像很难准...目的相比于一般视频目标检测跟踪任务,视频羽毛球的实时定位跟踪主要面临两大难点:1)羽毛球属于小目标,同时伴有严重的运动模糊以及相似目标的干扰,使用基于矩形框的目标检测跟踪方法准确率低且会带来中心点误差问题;2)单帧图像很难准确定位羽毛球目标,利用视频前后帧的时域特征则可以跟踪到羽毛球目标,而现有提取时域特征的网络模块结构复杂,难以满足实时性要求。针对以上问题,本文使用热力图轮廓检测方法,提出了羽毛球运动小目标的定位跟踪网络算法(shuttlecock track net,STNet)。方法网络主体采用“U”型编解码结构;针对小目标像素信息少的问题,基于SE(squeeze and excitation)通道注意力与残差结构设计高效特征提取模块(SE channel attention and residual,SECAR),实现了空域信息的高效提取与传递,提高了网络的定位性能;针对目标丢失与相似目标干扰问题,设计了时序网络(temporal network,TPN)结构用于提取和记忆视频时域特征,提高了网络跟踪性能。结果在羽毛球比赛公开数据集TrackNetv2与自制数据集上的实验表明,本文方法在多个指标上取得了最好的性能表现。相较于现有性能较好的羽毛球定位跟踪方法TrackNetv2,本文方法在准确率、精确率和F1上分别提高7.5%、15.7%和7.5%,并且显著降低了参数量,满足实时处理需求(54帧/s)。结论本文提出的STNet羽毛球定位跟踪网络,在面对羽毛球目标外观剧烈变化以及背景干扰严重时,能够准确定位羽毛球比赛视频帧中可能存在的羽毛球,实现羽毛球的稳定跟踪,相比其他羽毛球定位跟踪网络,具有更优的性能。展开更多
为探究不同加工处理下李光杏脯的理化特性和挥发性成分之间的差异,对传统工艺杏脯(hot air dried apricot,HAD)、复合糖液糖煮杏脯(boiled sugar with apricot,BSW)、弱微波处理杏脯(weak microwave drying of apricot,WMD)进行感官评...为探究不同加工处理下李光杏脯的理化特性和挥发性成分之间的差异,对传统工艺杏脯(hot air dried apricot,HAD)、复合糖液糖煮杏脯(boiled sugar with apricot,BSW)、弱微波处理杏脯(weak microwave drying of apricot,WMD)进行感官评定、褐变度、总糖、可滴定酸等的测定,采用顶空固相微萃取和气相色谱-质谱联用(headspace solid-phase microextraction-gas chromatography-mass spectrometry,HS-SPME-GC-MS)技术进行挥发性成分测定。研究结果表明,HAD的褐变度最高为0.66,BSW的褐变度最低为0.26。WMD处理的总糖含量最低为42.36%,可滴定酸含量为1.03%,WMD可有效降低含糖量,缓解果脯产品口感甜腻的问题。HS-SPME-GC-MS共检测出88种挥发性成分,主要包括醇类19种、醛类20种、酮类13种、酯类9种、萜烯类16种、酸性物质3种以及其他类8种。采用气味活度值结合正交偏最小二乘判别分析确定了3种样品的关键差异香气成分。其中HAD提取的关键香气成分为呋喃、联苯、3-糠醛;BSW的关键香气成分为芳樟醇、己醛、顺式-香叶醇;WMD的关键香气成分为己醛、芳樟醇、3-辛醇。该文通过对3种不同处理工艺的杏脯进行理化指标测定,结合与鲜果挥发性成分对比研究,为进一步阐明不同处理杏脯的理化性质和风味物质差异研究及合理开发相关产品提供科学基础。展开更多
基金supported by the Key Project of the NationalLanguage Commission(No.ZDI145-110)the AcademicResearch Projects of Beijing Union University(No.ZK20202514)+1 种基金the Key Laboratory Project(No.YYZN-2024-6)the Project for the Construction and Support of High-Level Innovative Teams in Beijing Municipal Institutions(No.BPHR20220121).
文摘Gaze estimation,a crucial non-verbal communication cue,has achieved remarkable progress through convolutional neural networks.However,accurate gaze prediction in uncon-strained environments,particularly in extreme head poses,partial occlusions,and abnormal lighting,remains challenging.Existing models often struggle to effectively focus on discriminative ocular features,leading to suboptimal performance.To address these limitations,this paper proposes dual-branch gaze estimation with Gaussian mixture distribution heatmaps and dynamic adaptive loss function(DMGDL),a novel dual-branch gaze estimation algorithm.By introducing Gaussian mixture distribution heatmaps centered on pupil positions as spatial attention guides,the model is enabled to prioritize ocular regions.Additionally,a dual-branch network architecture is designed to separately extract features for yaw and pitch angles,enhancing flexibility and mitigating cross-angle interference.A dynamic adaptive loss function is further formulated to address discontinuities in angle estimation,improving robustness and convergence stability.Experimental evaluations on three benchmark datasets demonstrate that DMGDL outperforms state-of-the-art methods,achiev-ing a mean angular error of 3.98°on the Max-Planck institute for informatics face gaze(MPI-IFaceGaze)dataset,10.21°on the physically unconstrained gaze estimation in the wild(Gaze360)dataset and 6.14°on the real-time eye gaze estimation in natural environments(RT-Gene)dataset,exhibiting superior generalization and robustness.
基金financially supported by the National Basic Research Program of China(2011CB505300,2011CB505303)the National Natural Science Foundation of China(81603258,81673599,81773882)+4 种基金Key Research Project in Basic Science of Jiangsu College and University(14KJA360001)Youth Talent Project Funded by Shaanxi Higher Education Association for Science and Technology(20180307)333 High Level Talents Training Project Funded by Jiangsu Province(BRA2016387)financially supported by a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD)the Discipline Innovation Team Program of Shaanxi University of Chinese Medicine(2019-YL10)
文摘Although compatibility is highly advocated in traditional Chinese medicine(TCM),inappropriate combination of some herbs may reduce the therapeutic action and even produce toxic effects.Kansui and licorice,one of TCM“Eighteen Incompatible Medicaments”,are the most representative cases of improper herbal combination,which may still be applied simultaneously under given conditions.However,the potential mechanism of their compatibility and incompatibility is unclear.In the present study,two different ratios of kansui and licorice,representing their compatibility and incompatibility respectively,were designed to elucidate their interaction by comparative plasma/tissue metabolomics and a heatmap with relative fold change.As a result,glycocholic acid,prostaglandin F2a,dihydroceramide and sphinganine were screened out as the principal alternative biomarkers of compatibility group;sphinganine,dihydroceramide,arachidonic acid,leukotriene B4,acetoacetic acid and linoleic acid were those of incompatibility group.Based on the values of biomarkers in each tissue,the liver was identified as the compatible target organ,while the heart,liver,and kidney were the incompatible target organs.Furthermore,important pathways for compatibility and incompatibility were also constructed.These results help us to better understand and utilize the two herbs,and the study was the first to reveal some innate characters of herbs related to TCM“Eighteen Incompatible Medicaments”.
基金the National Natural Science Foundation of China under Grant No.62072255.
文摘Recognition of human gesture actions is a challenging issue due to the complex patterns in both visual andskeletal features. Existing gesture action recognition (GAR) methods typically analyze visual and skeletal data,failing to meet the demands of various scenarios. Furthermore, multi-modal approaches lack the versatility toefficiently process both uniformand disparate input patterns.Thus, in this paper, an attention-enhanced pseudo-3Dresidual model is proposed to address the GAR problem, called HgaNets. This model comprises two independentcomponents designed formodeling visual RGB (red, green and blue) images and 3Dskeletal heatmaps, respectively.More specifically, each component consists of two main parts: 1) a multi-dimensional attention module forcapturing important spatial, temporal and feature information in human gestures;2) a spatiotemporal convolutionmodule that utilizes pseudo-3D residual convolution to characterize spatiotemporal features of gestures. Then,the output weights of the two components are fused to generate the recognition results. Finally, we conductedexperiments on four datasets to assess the efficiency of the proposed model. The results show that the accuracy onfour datasets reaches 85.40%, 91.91%, 94.70%, and 95.30%, respectively, as well as the inference time is 0.54 s andthe parameters is 2.74M. These findings highlight that the proposed model outperforms other existing approachesin terms of recognition accuracy.
文摘Musculoskeletal injuries are among the most common causes of disability worldwide,with early detection and appropriate intervention critical to minimizing long-term complications.Infrared thermography(IRT)has emerged as a noninvasive,real-time imaging modality that captures superficial temperature changes reflecting underlying physiological processes such as inflammation and vascular alterations.This review explores the fundamental principles of medical thermography,differentiates between passive and active approaches,and outlines key technological advancements including artificial intelligence integration.The clinical utility of IRT is discussed in various contexts–ranging from acute soft tissue injuries and overuse syndromes to chronic pain and rehabilitation monitoring.Comparative insights with conventional imaging techniques such as ultrasound and magnetic resonance imaging are also presented.While IRT offers functional imaging capabilities with advantages in portability,safety,and speed,its limitations–such as lack of deep-tissue penetration and protocol standardization–remain significant barriers to broader adoption.Future directions include the integration of IRT with other imaging modalities and digital health platforms to enhance musculoskeletal assessment and injury prevention strategies.
基金supported by the National Key Research and Devel-opment Project of China(2021YFD1600101)the earmarked fund for the China Agriculture Research System(CARS-12,CARS-13 and CARS-14)the Agricultural Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences(CAAS-ASTIP-2024-OCRI).
文摘Flax(Linum usitatissimum L.)is an important oil crop in the high-altitude arid regions of China.Flaxseed is rich in various nutrients.However,the nutritional qualities of flaxseeds from different producing areas are still unclear.In this study,the nutritional characteristics of flaxseed from five producing areas in China were investigated.Twenty five nutritional quality indices in flaxseed were analyzed.Subsequently,chemometric methods,including cluster analysis,principal component analysis(PCA)and partial least square discriminant analysis(PLS-DA),were employed to discover the characteristics of nutritional qualities in flaxseeds.The results revealed there are significant differences in nutritional qualities among flaxseeds from different production areas.Six quality indices includingγ-tocopherol,vitamin E,phytosterols,oleic acid,α-linolenic acid,and cycloartenol were susceptible to producing area.In detail,the superiorcharacteristic nutrients of Ningxia flaxseed,Inner Mongolia flaxseed and Hebei flaxseed are vitamin E(17.3 mg/100g),α-linolenic acid(52.6%)and cycloartenol(1738.1 mg/kg),and phytosterols(3032.0 mg/kg),respectively.This study promotes the high-value development and utilization of local flaxseed industry.
文摘目的相比于一般视频目标检测跟踪任务,视频羽毛球的实时定位跟踪主要面临两大难点:1)羽毛球属于小目标,同时伴有严重的运动模糊以及相似目标的干扰,使用基于矩形框的目标检测跟踪方法准确率低且会带来中心点误差问题;2)单帧图像很难准确定位羽毛球目标,利用视频前后帧的时域特征则可以跟踪到羽毛球目标,而现有提取时域特征的网络模块结构复杂,难以满足实时性要求。针对以上问题,本文使用热力图轮廓检测方法,提出了羽毛球运动小目标的定位跟踪网络算法(shuttlecock track net,STNet)。方法网络主体采用“U”型编解码结构;针对小目标像素信息少的问题,基于SE(squeeze and excitation)通道注意力与残差结构设计高效特征提取模块(SE channel attention and residual,SECAR),实现了空域信息的高效提取与传递,提高了网络的定位性能;针对目标丢失与相似目标干扰问题,设计了时序网络(temporal network,TPN)结构用于提取和记忆视频时域特征,提高了网络跟踪性能。结果在羽毛球比赛公开数据集TrackNetv2与自制数据集上的实验表明,本文方法在多个指标上取得了最好的性能表现。相较于现有性能较好的羽毛球定位跟踪方法TrackNetv2,本文方法在准确率、精确率和F1上分别提高7.5%、15.7%和7.5%,并且显著降低了参数量,满足实时处理需求(54帧/s)。结论本文提出的STNet羽毛球定位跟踪网络,在面对羽毛球目标外观剧烈变化以及背景干扰严重时,能够准确定位羽毛球比赛视频帧中可能存在的羽毛球,实现羽毛球的稳定跟踪,相比其他羽毛球定位跟踪网络,具有更优的性能。
文摘为探究不同加工处理下李光杏脯的理化特性和挥发性成分之间的差异,对传统工艺杏脯(hot air dried apricot,HAD)、复合糖液糖煮杏脯(boiled sugar with apricot,BSW)、弱微波处理杏脯(weak microwave drying of apricot,WMD)进行感官评定、褐变度、总糖、可滴定酸等的测定,采用顶空固相微萃取和气相色谱-质谱联用(headspace solid-phase microextraction-gas chromatography-mass spectrometry,HS-SPME-GC-MS)技术进行挥发性成分测定。研究结果表明,HAD的褐变度最高为0.66,BSW的褐变度最低为0.26。WMD处理的总糖含量最低为42.36%,可滴定酸含量为1.03%,WMD可有效降低含糖量,缓解果脯产品口感甜腻的问题。HS-SPME-GC-MS共检测出88种挥发性成分,主要包括醇类19种、醛类20种、酮类13种、酯类9种、萜烯类16种、酸性物质3种以及其他类8种。采用气味活度值结合正交偏最小二乘判别分析确定了3种样品的关键差异香气成分。其中HAD提取的关键香气成分为呋喃、联苯、3-糠醛;BSW的关键香气成分为芳樟醇、己醛、顺式-香叶醇;WMD的关键香气成分为己醛、芳樟醇、3-辛醇。该文通过对3种不同处理工艺的杏脯进行理化指标测定,结合与鲜果挥发性成分对比研究,为进一步阐明不同处理杏脯的理化性质和风味物质差异研究及合理开发相关产品提供科学基础。
