To enhance the efficiency and accuracy of environmental perception for autonomous vehicles,we propose GDMNet,a unified multi-task perception network for autonomous driving,capable of performing drivable area segmentat...To enhance the efficiency and accuracy of environmental perception for autonomous vehicles,we propose GDMNet,a unified multi-task perception network for autonomous driving,capable of performing drivable area segmentation,lane detection,and traffic object detection.Firstly,in the encoding stage,features are extracted,and Generalized Efficient Layer Aggregation Network(GELAN)is utilized to enhance feature extraction and gradient flow.Secondly,in the decoding stage,specialized detection heads are designed;the drivable area segmentation head employs DySample to expand feature maps,the lane detection head merges early-stage features and processes the output through the Focal Modulation Network(FMN).Lastly,the Minimum Point Distance IoU(MPDIoU)loss function is employed to compute the matching degree between traffic object detection boxes and predicted boxes,facilitating model training adjustments.Experimental results on the BDD100K dataset demonstrate that the proposed network achieves a drivable area segmentation mean intersection over union(mIoU)of 92.2%,lane detection accuracy and intersection over union(IoU)of 75.3%and 26.4%,respectively,and traffic object detection recall and mAP of 89.7%and 78.2%,respectively.The detection performance surpasses that of other single-task or multi-task algorithm models.展开更多
With the vigorous development of social economy in China,various advanced technologies and equipment have emerged,among which artificial intelligence(AI)has rapidly developed and achieved remarkable results when appli...With the vigorous development of social economy in China,various advanced technologies and equipment have emerged,among which artificial intelligence(AI)has rapidly developed and achieved remarkable results when applied to many fields.Therefore,leaders and teachers in primary and secondary schools should pay more attention to AI education and explore effective measures to optimize the effectiveness of this education.Among them,carrying out artificial intelligence education and teaching from the perspective of thinking quality,with an aim to improve students’technical ability and effectively cultivate their thinking skills,may improve students’learning efficiency and teachers’teaching efficiency.How to carry out AI education from the perspective of thinking quality is an important issue that teachers need to address urgently.Through in-depth research,we focus on this issue,in hope to benefit primary and secondary school teachers.展开更多
Uncovering mate choice and factors that lead to the choice are very important to understanding sexual selection in evolutionary change.Cicadas are known for their loud sounds produced by males using the timbals.Howeve...Uncovering mate choice and factors that lead to the choice are very important to understanding sexual selection in evolutionary change.Cicadas are known for their loud sounds produced by males using the timbals.However,males in certain cicada species emit 2 kinds of sounds using respectively timbals and stridulatory organs,and females may produce their own sounds to respond to males.What has never been considered is the mate choice in such cicada species.Here,we investigate the sexual selection and potential impact of predation pressure on mate choice in the cicada Subpsaltria yangi Chen.It possesses stridulatory sound-producing organs in both sexes in addition to the timbals in males.Results show that males producing calling songs with shorter timbal–stridulatory sound intervals and a higher call rate achieved greater mating success.No morphological traits were found to be correlated with mating success in both sexes,suggesting neither males nor females display mate preference for the opposite sex based on morphological traits.Males do not discriminate among responding females during mate searching,which may be due to the high energy costs associated with their unusual mate-seeking activity and the male-biased predation pressure.Females generally mate once but a minority of them re-mated after oviposition which,combined with the desirable acoustic traits of males,suggest females may maximize their reproductive success by choosing a high-quality male in the first place.This study contributes to our understanding mechanisms of sexual selection in cicadas and other insects suffering selective pressure from predators.展开更多
文摘To enhance the efficiency and accuracy of environmental perception for autonomous vehicles,we propose GDMNet,a unified multi-task perception network for autonomous driving,capable of performing drivable area segmentation,lane detection,and traffic object detection.Firstly,in the encoding stage,features are extracted,and Generalized Efficient Layer Aggregation Network(GELAN)is utilized to enhance feature extraction and gradient flow.Secondly,in the decoding stage,specialized detection heads are designed;the drivable area segmentation head employs DySample to expand feature maps,the lane detection head merges early-stage features and processes the output through the Focal Modulation Network(FMN).Lastly,the Minimum Point Distance IoU(MPDIoU)loss function is employed to compute the matching degree between traffic object detection boxes and predicted boxes,facilitating model training adjustments.Experimental results on the BDD100K dataset demonstrate that the proposed network achieves a drivable area segmentation mean intersection over union(mIoU)of 92.2%,lane detection accuracy and intersection over union(IoU)of 75.3%and 26.4%,respectively,and traffic object detection recall and mAP of 89.7%and 78.2%,respectively.The detection performance surpasses that of other single-task or multi-task algorithm models.
基金supported by the 2021 Guangdong Province General Universities Special Project in Key Areas(New Generation Information Technology)“Research on Building a Education Knowledge Graph Model for Higher Vocational Construction Major Supported by Artificial Intelligence”(Project No.2021ZDZX1112)the 2022 Higher Education Research Project of Guangdong Higher Education Association’s“14th Five Year Plan”“Research and Practice on the Cooperative Development Path of Higher Education in the Guangdong Hong Kong Macao Greater Bay Area from the Perspective of Supply Side Reform”(Project No.22GYB161).
文摘With the vigorous development of social economy in China,various advanced technologies and equipment have emerged,among which artificial intelligence(AI)has rapidly developed and achieved remarkable results when applied to many fields.Therefore,leaders and teachers in primary and secondary schools should pay more attention to AI education and explore effective measures to optimize the effectiveness of this education.Among them,carrying out artificial intelligence education and teaching from the perspective of thinking quality,with an aim to improve students’technical ability and effectively cultivate their thinking skills,may improve students’learning efficiency and teachers’teaching efficiency.How to carry out AI education from the perspective of thinking quality is an important issue that teachers need to address urgently.Through in-depth research,we focus on this issue,in hope to benefit primary and secondary school teachers.
基金This study was funded by the National Natural Science Foundation of China(Grant Nos.31772505 and 32070476).
文摘Uncovering mate choice and factors that lead to the choice are very important to understanding sexual selection in evolutionary change.Cicadas are known for their loud sounds produced by males using the timbals.However,males in certain cicada species emit 2 kinds of sounds using respectively timbals and stridulatory organs,and females may produce their own sounds to respond to males.What has never been considered is the mate choice in such cicada species.Here,we investigate the sexual selection and potential impact of predation pressure on mate choice in the cicada Subpsaltria yangi Chen.It possesses stridulatory sound-producing organs in both sexes in addition to the timbals in males.Results show that males producing calling songs with shorter timbal–stridulatory sound intervals and a higher call rate achieved greater mating success.No morphological traits were found to be correlated with mating success in both sexes,suggesting neither males nor females display mate preference for the opposite sex based on morphological traits.Males do not discriminate among responding females during mate searching,which may be due to the high energy costs associated with their unusual mate-seeking activity and the male-biased predation pressure.Females generally mate once but a minority of them re-mated after oviposition which,combined with the desirable acoustic traits of males,suggest females may maximize their reproductive success by choosing a high-quality male in the first place.This study contributes to our understanding mechanisms of sexual selection in cicadas and other insects suffering selective pressure from predators.