An action recognition network that combines multi-level spatiotemporal feature fusion with an attention mechanism is proposed as a solution to the issues of single spatiotemporal feature scale extraction,information r...An action recognition network that combines multi-level spatiotemporal feature fusion with an attention mechanism is proposed as a solution to the issues of single spatiotemporal feature scale extraction,information redundancy,and insufficient extraction of frequency domain information in channels in 3D convolutional neural networks.Firstly,based on 3D CNN,this paper designs a new multilevel spatiotemporal feature fusion(MSF)structure,which is embedded in the network model,mainly through multilevel spatiotemporal feature separation,splicing and fusion,to achieve the fusion of spatial perceptual fields and short-medium-long time series information at different scales with reduced network parameters;In the second step,a multi-frequency channel and spatiotemporal attention module(FSAM)is introduced to assign different frequency features and spatiotemporal features in the channels are assigned corresponding weights to reduce the information redundancy of the feature maps.Finally,we embed the proposed method into the R3D model,which replaced the 2D convolutional filters in the 2D Resnet with 3D convolutional filters and conduct extensive experimental validation on the small and medium-sized dataset UCF101 and the largesized dataset Kinetics-400.The findings revealed that our model increased the recognition accuracy on both datasets.Results on the UCF101 dataset,in particular,demonstrate that our model outperforms R3D in terms of a maximum recognition accuracy improvement of 7.2%while using 34.2%fewer parameters.The MSF and FSAM are migrated to another traditional 3D action recognition model named C3D for application testing.The test results based on UCF101 show that the recognition accuracy is improved by 8.9%,proving the strong generalization ability and universality of the method in this paper.展开更多
选取2005年5月24日-6月18日在金塔开展的“绿洲系统能量与水分循环过程的观测实验”中的3层CSAT3的实验数据，应用Schmid的FSAM（The Flux-Source Area Model）模型，分析了不同观测高度的通量贡献源区分布以及观测高度对通量贡献源区...选取2005年5月24日-6月18日在金塔开展的“绿洲系统能量与水分循环过程的观测实验”中的3层CSAT3的实验数据，应用Schmid的FSAM（The Flux-Source Area Model）模型，分析了不同观测高度的通量贡献源区分布以及观测高度对通量贡献源区分布的影响，同时分析了不同大气层结条件下源区的分布以及稳定度对通量贡献源区分布的影响。结果表明,稳定条件下的通量贡献源区大于不稳定条件下的通量贡献源区，并且随着观测高度的增加通量贡献源区会显著增大。展开更多
[Objective] The paper was to analyze the quality of flux observation data of rubber plantation.[Method]Based on the FSAM model,footprint and flux source area were analyzed according to the continuous flux measurement ...[Objective] The paper was to analyze the quality of flux observation data of rubber plantation.[Method]Based on the FSAM model,footprint and flux source area were analyzed according to the continuous flux measurement with the open-path eddy covariance system on the 50 m tower of Danzhou Key Field Station of Observation and Research for Tropical Agricultural Resources and Environments,Ministry of Agriculture from Jan 1 to Jun 30,2010.[Result] Under unstable stratification,source areas were smaller than those under stable conditions,and source areas in the dormant season were larger than those in the growing season at the same level.In the main wind direction 130°-270°,the upwind range of source areas was in the magnitude of 100-758 m and vertical upwind range was-251-251 m at 80% level under unstable stratification in the growing season,and they were some large than those under the unstable stratification in the dormant season.The source areas of the upwind and vertical upwind ranges were 173-1 858,-534-534 m especially under stable stratification in the growing season,and they were smaller than those under stable stratification in the dormant season.In the other wind directions of 0°-130° and 270°-360°,the ranges were similar to those of the growing season in the prevailing wind direction under the same atmospheric conditions.[Conclusion] The study would lay a foundation for the future flux calculation and analysis.展开更多
基金supported by the General Program of the National Natural Science Foundation of China (62272234)the Enterprise Cooperation Project (2022h160)the Priority Academic Program Development of Jiangsu Higher Education Institutions Project.
文摘An action recognition network that combines multi-level spatiotemporal feature fusion with an attention mechanism is proposed as a solution to the issues of single spatiotemporal feature scale extraction,information redundancy,and insufficient extraction of frequency domain information in channels in 3D convolutional neural networks.Firstly,based on 3D CNN,this paper designs a new multilevel spatiotemporal feature fusion(MSF)structure,which is embedded in the network model,mainly through multilevel spatiotemporal feature separation,splicing and fusion,to achieve the fusion of spatial perceptual fields and short-medium-long time series information at different scales with reduced network parameters;In the second step,a multi-frequency channel and spatiotemporal attention module(FSAM)is introduced to assign different frequency features and spatiotemporal features in the channels are assigned corresponding weights to reduce the information redundancy of the feature maps.Finally,we embed the proposed method into the R3D model,which replaced the 2D convolutional filters in the 2D Resnet with 3D convolutional filters and conduct extensive experimental validation on the small and medium-sized dataset UCF101 and the largesized dataset Kinetics-400.The findings revealed that our model increased the recognition accuracy on both datasets.Results on the UCF101 dataset,in particular,demonstrate that our model outperforms R3D in terms of a maximum recognition accuracy improvement of 7.2%while using 34.2%fewer parameters.The MSF and FSAM are migrated to another traditional 3D action recognition model named C3D for application testing.The test results based on UCF101 show that the recognition accuracy is improved by 8.9%,proving the strong generalization ability and universality of the method in this paper.
文摘选取2005年5月24日-6月18日在金塔开展的“绿洲系统能量与水分循环过程的观测实验”中的3层CSAT3的实验数据，应用Schmid的FSAM（The Flux-Source Area Model）模型，分析了不同观测高度的通量贡献源区分布以及观测高度对通量贡献源区分布的影响，同时分析了不同大气层结条件下源区的分布以及稳定度对通量贡献源区分布的影响。结果表明,稳定条件下的通量贡献源区大于不稳定条件下的通量贡献源区，并且随着观测高度的增加通量贡献源区会显著增大。
基金Supported by the Fundamental Research Funds for Rubber Research Institute,CATAS (1630022011013 )Hainan Natural Science Foundation (807045)Running Costs of Danzhou Key Field Station of Observation and Research for Tropical Agricultural Resources and Environments,Ministry of Agriculture~~
文摘[Objective] The paper was to analyze the quality of flux observation data of rubber plantation.[Method]Based on the FSAM model,footprint and flux source area were analyzed according to the continuous flux measurement with the open-path eddy covariance system on the 50 m tower of Danzhou Key Field Station of Observation and Research for Tropical Agricultural Resources and Environments,Ministry of Agriculture from Jan 1 to Jun 30,2010.[Result] Under unstable stratification,source areas were smaller than those under stable conditions,and source areas in the dormant season were larger than those in the growing season at the same level.In the main wind direction 130°-270°,the upwind range of source areas was in the magnitude of 100-758 m and vertical upwind range was-251-251 m at 80% level under unstable stratification in the growing season,and they were some large than those under the unstable stratification in the dormant season.The source areas of the upwind and vertical upwind ranges were 173-1 858,-534-534 m especially under stable stratification in the growing season,and they were smaller than those under stable stratification in the dormant season.In the other wind directions of 0°-130° and 270°-360°,the ranges were similar to those of the growing season in the prevailing wind direction under the same atmospheric conditions.[Conclusion] The study would lay a foundation for the future flux calculation and analysis.
