自然场景中的实体标志,如商标、交通标志等,易受拍摄角度、所依附物体形变、尺度变化等影响,导致检测精度降低.为此,提出一种注意力引导的标志检测与识别网络(attention guided logo detection and recognition network,AGLDN),联合优...自然场景中的实体标志,如商标、交通标志等,易受拍摄角度、所依附物体形变、尺度变化等影响,导致检测精度降低.为此,提出一种注意力引导的标志检测与识别网络(attention guided logo detection and recognition network,AGLDN),联合优化模型对多尺度变化和复杂形变的鲁棒性.首先通过标志模板图像搜集及掩码生成、标志背景图像选取和标志图像生成创建标志合成数据集;然后基于RetinaNet和FPN提取多尺度特征并形成高级语义特征映射;最后利用注意力机制引导网络关注标志区域,克服目标变形对特征鲁棒性的影响,实现标志检测与识别.实验结果表明,所提方法可以有效降低尺度变化、非刚性形变的影响,提高标志检测准确率.展开更多
In recent years,the escalation in accidental explosions has emerged as a formidable threat to tunnel infrastructures.Therefore,it is of great significance to conduct a dynamic performance analysis of the tunnels,to im...In recent years,the escalation in accidental explosions has emerged as a formidable threat to tunnel infrastructures.Therefore,it is of great significance to conduct a dynamic performance analysis of the tunnels,to improve the safety and maintain the functionality of underground transport hubs.To this end,this study proposes a dynamic performance assessment framework to assess the extent of damage of shallow buried circular tunnels under explosion hazards.First,the nonlinear dynamic finite element numerical model of soil-tunnel interaction system under explosion hazard was established and validated.Then,based on the validated numerical model,an explosion intensity(EI)considering both explosion equivalent and relative distance was used to further analyze the dynamic response characteristics under typical explosion conditions.Finally,this study further explored the influence of the integrity and strength of the surrounding soil,concrete strength,lining thickness,rebar strength,and rebar rate on the tunnel dynamic performance.Our results show that the dynamic performance assessment framework proposed for shallow circular tunnels fully integrates the coupling effects of explosion equivalent and distance,and is able to accurately measure the degree of damage sustained by these structures under different EI.This work contributes to designing and managing tunnels and underground transport networks based on dynamic performance,thereby facilitating decision-making and efficient allocation of resources by consultants,operators,and stakeholders.展开更多
The rapid evolution of the digital economy has brought significant opportunities and challenges to the field of geotechnical and underground engineering.As the scale of these projects continues to grow,their planning,...The rapid evolution of the digital economy has brought significant opportunities and challenges to the field of geotechnical and underground engineering.As the scale of these projects continues to grow,their planning,design,construction,and maintenance generate extensive multi-source and heterogeneous data.These data include site investigation reports,design blueprints,and structural deformation monitoring records,all of which capture the intricate and diverse characteristics of geotechnical systems.Additionally,risks and hazards during the construction phase pose potential threats to engineering safety,enhancing the need for effective risk management strategies.展开更多
文摘自然场景中的实体标志,如商标、交通标志等,易受拍摄角度、所依附物体形变、尺度变化等影响,导致检测精度降低.为此,提出一种注意力引导的标志检测与识别网络(attention guided logo detection and recognition network,AGLDN),联合优化模型对多尺度变化和复杂形变的鲁棒性.首先通过标志模板图像搜集及掩码生成、标志背景图像选取和标志图像生成创建标志合成数据集;然后基于RetinaNet和FPN提取多尺度特征并形成高级语义特征映射;最后利用注意力机制引导网络关注标志区域,克服目标变形对特征鲁棒性的影响,实现标志检测与识别.实验结果表明,所提方法可以有效降低尺度变化、非刚性形变的影响,提高标志检测准确率.
基金Project(22dz1201202)supported by the Shanghai Science and Technology Committee Program,ChinaProjects(52108381,52090082)supported by the National Natural Science Foundation of China+1 种基金Project(2023QNRC001)supported by the Young Elite Scientists Sponsorship Program by CAST,ChinaProject(TSY2022QT161)supported by the Damage Database for Urban Rail Transit Underground Structures and Resilience Evaluation Algorithm Research。
文摘In recent years,the escalation in accidental explosions has emerged as a formidable threat to tunnel infrastructures.Therefore,it is of great significance to conduct a dynamic performance analysis of the tunnels,to improve the safety and maintain the functionality of underground transport hubs.To this end,this study proposes a dynamic performance assessment framework to assess the extent of damage of shallow buried circular tunnels under explosion hazards.First,the nonlinear dynamic finite element numerical model of soil-tunnel interaction system under explosion hazard was established and validated.Then,based on the validated numerical model,an explosion intensity(EI)considering both explosion equivalent and relative distance was used to further analyze the dynamic response characteristics under typical explosion conditions.Finally,this study further explored the influence of the integrity and strength of the surrounding soil,concrete strength,lining thickness,rebar strength,and rebar rate on the tunnel dynamic performance.Our results show that the dynamic performance assessment framework proposed for shallow circular tunnels fully integrates the coupling effects of explosion equivalent and distance,and is able to accurately measure the degree of damage sustained by these structures under different EI.This work contributes to designing and managing tunnels and underground transport networks based on dynamic performance,thereby facilitating decision-making and efficient allocation of resources by consultants,operators,and stakeholders.
文摘The rapid evolution of the digital economy has brought significant opportunities and challenges to the field of geotechnical and underground engineering.As the scale of these projects continues to grow,their planning,design,construction,and maintenance generate extensive multi-source and heterogeneous data.These data include site investigation reports,design blueprints,and structural deformation monitoring records,all of which capture the intricate and diverse characteristics of geotechnical systems.Additionally,risks and hazards during the construction phase pose potential threats to engineering safety,enhancing the need for effective risk management strategies.
