Polymer backbone plays a fundamental role in determining the molecular architecture and properties of polymers.Polymers can be modified by incorporating heteroatoms into their backbones to achieve tunable optical and ...Polymer backbone plays a fundamental role in determining the molecular architecture and properties of polymers.Polymers can be modified by incorporating heteroatoms into their backbones to achieve tunable optical and mechanical properties,such as polyamide,polythioether and polysilane[1].The transition from nonconjugated to conjugated backbones of polymers challenges the traditional view of polymers as insulators,leading to the development of conductive polymers[2].In contrast,metal elements far surpass non-metal elements in both the variety of elemental types and the diversity of their outer-shell electrons,incorporating which into the polymer backbone is promising to create polymer materials with unique properties and applications,such as high mechanical strength and electrical conductivity[3].Integration of metal atoms into the polymer backbones has been reported.However,the lack of uniformity and continuity in the interaction of metal atoms limits electron transfer efficiency and hinders the full utilization of metal elements within polymer materials[4].To this end,a novel metal-backboned polymer was proposed[5,6],wherein the polymer backbone consists entirely of metal atoms interconnected through metal–metal bonds.This novel polymer was found with exceptional optical and electrical properties,showing promising applications in photoelectric devices,flexible electronics,and microwave absorption materials[7,8].展开更多
Materials are generally categorized as three main types of metal,non-metal and organic polymer.Among them,metal and organic polymers show opposite advantages and disadvantages.For instance,a metal exhibits both high m...Materials are generally categorized as three main types of metal,non-metal and organic polymer.Among them,metal and organic polymers show opposite advantages and disadvantages.For instance,a metal exhibits both high mechanical strength and electrical conductivity while both low flexibility and solution processibility;an organic polymer exhibits both high flexibility and solution processibility but with both low mechanical strength and electrical conductivity[1–4].Although a lot of efforts are made to synthesize metal/polymer composite materials to partly maintain advantages of each component[5–7],people have never thought that it is possible to achieve the advantages of both metal and polymer by synthesizing novel material with a single component.展开更多
This paper proposes a lightweight traffic sign detection system based on you only look once(YOLO).Firstly,the classification to fusion(C2f)structure is integrated into the backbone network,employing deformable convolu...This paper proposes a lightweight traffic sign detection system based on you only look once(YOLO).Firstly,the classification to fusion(C2f)structure is integrated into the backbone network,employing deformable convolution and bi-directional feature pyramid network(BiFPN)_Concat to improve the adaptability of the network.Secondly,the simple attention module(SimAm)is embedded to prioritize key features and reduce the complexity of the model after the C2f layer at the end of the backbone network.Next,the focal efficient intersection over union(EloU)is introduced to adjust the weights of challenging samples.Finally,we accomplish the design and deployment for the mobile app.The results demonstrate improvements,with the F1 score of 0.8987,mean average precision(mAP)@0.5 of 98.8%,mAP@0.5:0.95 of 75.6%,and the detection speed of 50 frames per second(FPS).展开更多
Aiming at the problem that the current traffic safety helmet detection model can't balance the accuracy of detection with the size of the model and the poor generalization of the model,a method based on improving ...Aiming at the problem that the current traffic safety helmet detection model can't balance the accuracy of detection with the size of the model and the poor generalization of the model,a method based on improving you only look once version 5(YOLOv5) is proposed.By incorporating the lightweight Ghost Net module into the YOLOv5 backbone network,we effectively reduce the model size.The addition of the receptive fields block(RFB) module enhances feature extraction and improves the feature acquisition capability of the lightweight model.Subsequently,the high-performance lightweight convolution,GSConv,is integrated into the neck structure for further model size compression.Moreover,the baseline model's loss function is substituted with efficient insertion over union(EIoU),accelerating network convergence and enhancing detection precision.Experimental results corroborate the effectiveness of this improved algorithm in real-world traffic scenarios.展开更多
Chain-growth radical polymerization of vinyl monomers is essential for producing a wide range of materials with properties tailored to specific applications.However,the inherent resistance of the polymer's C―C ba...Chain-growth radical polymerization of vinyl monomers is essential for producing a wide range of materials with properties tailored to specific applications.However,the inherent resistance of the polymer's C―C backbone to degradation raises significant concerns regarding long-term environmental persistence,which also limits their potential in biomedical applications.To address these challenges,researchers have developed strategies to either degrade preexisting vinyl polymers or incorporate cleavable units into the backbone to modify them with enhanced degradability.This review explores the various approaches aimed at achieving backbone degradability in chain-growth radical polymerization of vinyl monomers,while also highlighting future research directions for the development of application-driven degradable vinyl polymers.展开更多
Incorporating a low density of ester units into the backbone of polyethylene materials enhances their sustainability and recyclability while maintaining the main material properties of polyethylenes.Here we report a n...Incorporating a low density of ester units into the backbone of polyethylene materials enhances their sustainability and recyclability while maintaining the main material properties of polyethylenes.Here we report a new way to access degradable polyethylene materials with a low content of in-chain ester units via mechanochemical backbone editing.Initially,ester groups are incorporated as side groups through catalytic copolymerization of ethylene with a cyclobutene-fused lactone monomer(CBL),yielding polyethylene materials with high molecular weights and adjustable thermomechanical properties.Subsequent solid-state ball-milling treatment selectively introduces side-chain ester groups into the main chain of the polyethylene materials via force-induced cycloreversion of the cyclobutane units.Under acidic conditions,hydrolysis of the resultant polyethylene materials with in-chain ester units facilitates further degradation into oligomers.展开更多
This article discusses the detailed examination of the engineering design and implementation process for direct Train-to-Train(T2T)communication within a wireless train backbone network in the context of a virtual cou...This article discusses the detailed examination of the engineering design and implementation process for direct Train-to-Train(T2T)communication within a wireless train backbone network in the context of a virtual coupling scenario.The article proposed several critical aspects,including the optimization of transmission data requirements,which is essential to ensure that communication between trains is efficient and reliable.The design of the T2T wireless communication subsystem is discussed in detail,outlining the technical specifications,protocols,and technologies employed to facilitate wireless communication between multiple trains.Additionally,the article presents a thorough analysis of the data collected during real-world train experiments,highlighting the performance metrics and challenges encountered during testing.This empirical data not only validates the effectiveness of the proposed design but also serves as a crucial reference for future advancements in T2T wireless communication systems.By combining both theoretical principles and practical outcomes,the article offers insights that will aid engineers and researchers in developing robust and efficient wireless communication systems for next-generation train operations.展开更多
Since the introduction of vision Transformers into the computer vision field,many vision tasks such as semantic segmentation tasks,have undergone radical changes.Although Transformer enhances the correlation of each l...Since the introduction of vision Transformers into the computer vision field,many vision tasks such as semantic segmentation tasks,have undergone radical changes.Although Transformer enhances the correlation of each local feature of an image object in the hidden space through the attention mechanism,it is difficult for a segmentation head to accomplish the mask prediction for dense embedding of multi-category and multi-local features.We present patch prototype vision Transformer(PPFormer),a Transformer architecture for semantic segmentation based on knowledge-embedded patch prototypes.1)The hierarchical Transformer encoder can generate multi-scale and multi-layered patch features including seamless patch projection to obtain information of multiscale patches,and feature-clustered self-attention to enhance the interplay of multi-layered visual information with implicit position encodes.2)PPFormer utilizes a non-parametric prototype decoder to extract region observations which represent significant parts of the objects by unlearnable patch prototypes and then calculate similarity between patch prototypes and pixel embeddings.The proposed contrasting patch prototype alignment module,which uses new patch prototypes to update prototype bank,effectively maintains class boundaries for prototypes.For different application scenarios,we have launched PPFormer-S,PPFormer-M and PPFormer-L by expanding the scale.Experimental results demonstrate that PPFormer can outperform fully convolutional networks(FCN)-and attention-based semantic segmentation models on the PASCAL VOC 2012,ADE20k,and Cityscapes datasets.展开更多
Aryl-ether bonds are facile to attack by oxidizing radicals,thus stimulating the exploitation of ether-free polymers as proton exchange membranes(PEMs)for the long-lasting operation of fuel cells.In this study,a novel...Aryl-ether bonds are facile to attack by oxidizing radicals,thus stimulating the exploitation of ether-free polymers as proton exchange membranes(PEMs)for the long-lasting operation of fuel cells.In this study,a novel class of PEMs derived from all-carbon fluorinated backbone polymers containing sulfide-linked alkyl sulfonic acid side chains have been developed through a straightforward and effective synthetic procedure.The sulfide-linked alkyl sulfonate groups were tethered to the poly(triphenylene pentafluorophenyl)backbone through a quantified and site-specific para-fluoro-thiol click reaction.Owing to the existence of obvious phase separation morphology between hydrophobic main chain and hydrophilic sulfonate groups in the side chains,resulting PEMs demonstrated favorable proton conductivity of 142.5m S/cm at 80℃,while maintaining excellent dimensional stability with an in-plane swelling ratio of<17%as well as a through-plane swelling ratio of<25%.They also exhibit elevated thermal decomposition temperatures(Td5%exceeding 300℃)alongside high tensile strength(>50 MPa).Furthermore,the ether-free full-carbon fluorinated main chain and the-S-group in the side chain,which serves as an effective freeradical scavenger,providing good chemical stability during Fenton’s test.The PEMs achieved a maximum power density of 407 m W/cm^(2)in a single H^(2)/air fuel cell,and an open-circuit voltage decline rate of 0.275 m V/h in a durability test at 30%RH and 80℃.Concurrently,the hydrogen crossover current density is only 1/3 of that of Nafion 212.These findings reveal that the resulted PEMs display considerable antioxidative properties along with commendable performance,with prospective applications in proton exchange membrane fuel cells.展开更多
Dear Editor,Environmental pollution from microplastics(MPs)has recently gained attention as a potential environmental hazard(Chia et al.,2021).Agricultural soils could contain more MPs than the ocean by 2050 because m...Dear Editor,Environmental pollution from microplastics(MPs)has recently gained attention as a potential environmental hazard(Chia et al.,2021).Agricultural soils could contain more MPs than the ocean by 2050 because more MPs enter the soil than the ocean(Nizzetto et al.,2016).The carbon(C)-C backbone of degradation-resistant MPs provides considerable stability in the soil,where they can remain for several decades(Iqbal et al.,2023).展开更多
This paper proposes a novel method for the automatic diagnosis of keratitis using feature vector quantization and self-attention mechanisms(ADK_FVQSAM).First,high-level features are extracted using the DenseNet121 bac...This paper proposes a novel method for the automatic diagnosis of keratitis using feature vector quantization and self-attention mechanisms(ADK_FVQSAM).First,high-level features are extracted using the DenseNet121 backbone network,followed by adaptive average pooling to scale the features to a fixed length.Subsequently,product quantization with residuals(PQR)is applied to convert continuous feature vectors into discrete features representations,preserving essential information insensitive to image quality variations.The quantized and original features are concatenated and fed into a self-attention mechanism to capture keratitis-related features.Finally,these enhanced features are classified through a fully connected layer.Experiments on clinical low-quality(LQ)images show that ADK_FVQSAM achieves accuracies of 87.7%,81.9%,and 89.3% for keratitis,other corneal abnormalities,and normal corneas,respectively.Compared to DenseNet121,Swin transformer,and InceptionResNet,ADK_FVQSAM improves average accuracy by 3.1%,11.3%,and 15.3%,respectively.These results demonstrate that ADK_FVQSAM significantly enhances the recognition performance of keratitis based on LQ slit-lamp images,offering a practical approach for clinical application.展开更多
Aiming at the problems of low detection accuracy and large model size of existing object detection algorithms applied to complex road scenes,an improved you only look once version 8(YOLOv8)object detection algorithm f...Aiming at the problems of low detection accuracy and large model size of existing object detection algorithms applied to complex road scenes,an improved you only look once version 8(YOLOv8)object detection algorithm for infrared images,F-YOLOv8,is proposed.First,a spatial-to-depth network replaces the traditional backbone network's strided convolution or pooling layer.At the same time,it combines with the channel attention mechanism so that the neural network focuses on the channels with large weight values to better extract low-resolution image feature information;then an improved feature pyramid network of lightweight bidirectional feature pyramid network(L-BiFPN)is proposed,which can efficiently fuse features of different scales.In addition,a loss function of insertion of union based on the minimum point distance(MPDIoU)is introduced for bounding box regression,which obtains faster convergence speed and more accurate regression results.Experimental results on the FLIR dataset show that the improved algorithm can accurately detect infrared road targets in real time with 3%and 2.2%enhancement in mean average precision at 50%IoU(mAP50)and mean average precision at 50%—95%IoU(mAP50-95),respectively,and 38.1%,37.3%and 16.9%reduction in the number of model parameters,the model weight,and floating-point operations per second(FLOPs),respectively.To further demonstrate the detection capability of the improved algorithm,it is tested on the public dataset PASCAL VOC,and the results show that F-YOLO has excellent generalized detection performance.展开更多
基金supported by the Ministry of Science and Technology of the People’s Republic of China(2022YFA1203001 and 2022YFA1203002)the National Natural Science Foundation of China(T2321003,22335003,and 22105045)Science and Technology Commission of Shanghai Municipality(21511104900 and 20JC1414902).
文摘Polymer backbone plays a fundamental role in determining the molecular architecture and properties of polymers.Polymers can be modified by incorporating heteroatoms into their backbones to achieve tunable optical and mechanical properties,such as polyamide,polythioether and polysilane[1].The transition from nonconjugated to conjugated backbones of polymers challenges the traditional view of polymers as insulators,leading to the development of conductive polymers[2].In contrast,metal elements far surpass non-metal elements in both the variety of elemental types and the diversity of their outer-shell electrons,incorporating which into the polymer backbone is promising to create polymer materials with unique properties and applications,such as high mechanical strength and electrical conductivity[3].Integration of metal atoms into the polymer backbones has been reported.However,the lack of uniformity and continuity in the interaction of metal atoms limits electron transfer efficiency and hinders the full utilization of metal elements within polymer materials[4].To this end,a novel metal-backboned polymer was proposed[5,6],wherein the polymer backbone consists entirely of metal atoms interconnected through metal–metal bonds.This novel polymer was found with exceptional optical and electrical properties,showing promising applications in photoelectric devices,flexible electronics,and microwave absorption materials[7,8].
基金supported by Ministry of Science and Technology of the People’s Republic of China(MOST,2022YFA1203001 and 2022YFA1203002)National Natural Science Foundation of China(22105045,T2321003,22335003)Science and Technology Commission of Shanghai Municipality(21511104900)。
文摘Materials are generally categorized as three main types of metal,non-metal and organic polymer.Among them,metal and organic polymers show opposite advantages and disadvantages.For instance,a metal exhibits both high mechanical strength and electrical conductivity while both low flexibility and solution processibility;an organic polymer exhibits both high flexibility and solution processibility but with both low mechanical strength and electrical conductivity[1–4].Although a lot of efforts are made to synthesize metal/polymer composite materials to partly maintain advantages of each component[5–7],people have never thought that it is possible to achieve the advantages of both metal and polymer by synthesizing novel material with a single component.
基金supported by the Shanxi Agricultural University Science and Technology Innovation Enhancement Project。
文摘This paper proposes a lightweight traffic sign detection system based on you only look once(YOLO).Firstly,the classification to fusion(C2f)structure is integrated into the backbone network,employing deformable convolution and bi-directional feature pyramid network(BiFPN)_Concat to improve the adaptability of the network.Secondly,the simple attention module(SimAm)is embedded to prioritize key features and reduce the complexity of the model after the C2f layer at the end of the backbone network.Next,the focal efficient intersection over union(EloU)is introduced to adjust the weights of challenging samples.Finally,we accomplish the design and deployment for the mobile app.The results demonstrate improvements,with the F1 score of 0.8987,mean average precision(mAP)@0.5 of 98.8%,mAP@0.5:0.95 of 75.6%,and the detection speed of 50 frames per second(FPS).
文摘Aiming at the problem that the current traffic safety helmet detection model can't balance the accuracy of detection with the size of the model and the poor generalization of the model,a method based on improving you only look once version 5(YOLOv5) is proposed.By incorporating the lightweight Ghost Net module into the YOLOv5 backbone network,we effectively reduce the model size.The addition of the receptive fields block(RFB) module enhances feature extraction and improves the feature acquisition capability of the lightweight model.Subsequently,the high-performance lightweight convolution,GSConv,is integrated into the neck structure for further model size compression.Moreover,the baseline model's loss function is substituted with efficient insertion over union(EIoU),accelerating network convergence and enhancing detection precision.Experimental results corroborate the effectiveness of this improved algorithm in real-world traffic scenarios.
基金funding from the National Natural Science Foundation of China(No.22401037)funding from JST CREST(No.JPMJCR23L1)。
文摘Chain-growth radical polymerization of vinyl monomers is essential for producing a wide range of materials with properties tailored to specific applications.However,the inherent resistance of the polymer's C―C backbone to degradation raises significant concerns regarding long-term environmental persistence,which also limits their potential in biomedical applications.To address these challenges,researchers have developed strategies to either degrade preexisting vinyl polymers or incorporate cleavable units into the backbone to modify them with enhanced degradability.This review explores the various approaches aimed at achieving backbone degradability in chain-growth radical polymerization of vinyl monomers,while also highlighting future research directions for the development of application-driven degradable vinyl polymers.
基金financially supported by the National Natural Science Foundation of China(No.52473097)the Fundamental Research Funds for the Central Universities(No.24X010301678)Shanghai Jiao Tong University 2030 Initiative(No.WH510363002/002)。
文摘Incorporating a low density of ester units into the backbone of polyethylene materials enhances their sustainability and recyclability while maintaining the main material properties of polyethylenes.Here we report a new way to access degradable polyethylene materials with a low content of in-chain ester units via mechanochemical backbone editing.Initially,ester groups are incorporated as side groups through catalytic copolymerization of ethylene with a cyclobutene-fused lactone monomer(CBL),yielding polyethylene materials with high molecular weights and adjustable thermomechanical properties.Subsequent solid-state ball-milling treatment selectively introduces side-chain ester groups into the main chain of the polyethylene materials via force-induced cycloreversion of the cyclobutane units.Under acidic conditions,hydrolysis of the resultant polyethylene materials with in-chain ester units facilitates further degradation into oligomers.
基金supported by the National Key R&D Program of China(2021YFF0501103).
文摘This article discusses the detailed examination of the engineering design and implementation process for direct Train-to-Train(T2T)communication within a wireless train backbone network in the context of a virtual coupling scenario.The article proposed several critical aspects,including the optimization of transmission data requirements,which is essential to ensure that communication between trains is efficient and reliable.The design of the T2T wireless communication subsystem is discussed in detail,outlining the technical specifications,protocols,and technologies employed to facilitate wireless communication between multiple trains.Additionally,the article presents a thorough analysis of the data collected during real-world train experiments,highlighting the performance metrics and challenges encountered during testing.This empirical data not only validates the effectiveness of the proposed design but also serves as a crucial reference for future advancements in T2T wireless communication systems.By combining both theoretical principles and practical outcomes,the article offers insights that will aid engineers and researchers in developing robust and efficient wireless communication systems for next-generation train operations.
基金supported in part by the Gansu Haizhi Characteristic Demonstration Project(No.GSHZTS2022-2).
文摘Since the introduction of vision Transformers into the computer vision field,many vision tasks such as semantic segmentation tasks,have undergone radical changes.Although Transformer enhances the correlation of each local feature of an image object in the hidden space through the attention mechanism,it is difficult for a segmentation head to accomplish the mask prediction for dense embedding of multi-category and multi-local features.We present patch prototype vision Transformer(PPFormer),a Transformer architecture for semantic segmentation based on knowledge-embedded patch prototypes.1)The hierarchical Transformer encoder can generate multi-scale and multi-layered patch features including seamless patch projection to obtain information of multiscale patches,and feature-clustered self-attention to enhance the interplay of multi-layered visual information with implicit position encodes.2)PPFormer utilizes a non-parametric prototype decoder to extract region observations which represent significant parts of the objects by unlearnable patch prototypes and then calculate similarity between patch prototypes and pixel embeddings.The proposed contrasting patch prototype alignment module,which uses new patch prototypes to update prototype bank,effectively maintains class boundaries for prototypes.For different application scenarios,we have launched PPFormer-S,PPFormer-M and PPFormer-L by expanding the scale.Experimental results demonstrate that PPFormer can outperform fully convolutional networks(FCN)-and attention-based semantic segmentation models on the PASCAL VOC 2012,ADE20k,and Cityscapes datasets.
基金supported by the Development of Scientific and Technological Project of Jilin Province(No.20230201139GX)。
文摘Aryl-ether bonds are facile to attack by oxidizing radicals,thus stimulating the exploitation of ether-free polymers as proton exchange membranes(PEMs)for the long-lasting operation of fuel cells.In this study,a novel class of PEMs derived from all-carbon fluorinated backbone polymers containing sulfide-linked alkyl sulfonic acid side chains have been developed through a straightforward and effective synthetic procedure.The sulfide-linked alkyl sulfonate groups were tethered to the poly(triphenylene pentafluorophenyl)backbone through a quantified and site-specific para-fluoro-thiol click reaction.Owing to the existence of obvious phase separation morphology between hydrophobic main chain and hydrophilic sulfonate groups in the side chains,resulting PEMs demonstrated favorable proton conductivity of 142.5m S/cm at 80℃,while maintaining excellent dimensional stability with an in-plane swelling ratio of<17%as well as a through-plane swelling ratio of<25%.They also exhibit elevated thermal decomposition temperatures(Td5%exceeding 300℃)alongside high tensile strength(>50 MPa).Furthermore,the ether-free full-carbon fluorinated main chain and the-S-group in the side chain,which serves as an effective freeradical scavenger,providing good chemical stability during Fenton’s test.The PEMs achieved a maximum power density of 407 m W/cm^(2)in a single H^(2)/air fuel cell,and an open-circuit voltage decline rate of 0.275 m V/h in a durability test at 30%RH and 80℃.Concurrently,the hydrogen crossover current density is only 1/3 of that of Nafion 212.These findings reveal that the resulted PEMs display considerable antioxidative properties along with commendable performance,with prospective applications in proton exchange membrane fuel cells.
基金supported by the National Natural Science Foundation of China(Nos.42207353 and 42277408)the Key Research and Development Program of Jiangsu Province of China(BE2021378)+1 种基金Jiangsu Agricultural Science and Technology Independent Innovation Fund of China(CX(21)-1009)the Earmarked Fund of China Agriculture Research System(CARS-10-Sweetpotato)。
文摘Dear Editor,Environmental pollution from microplastics(MPs)has recently gained attention as a potential environmental hazard(Chia et al.,2021).Agricultural soils could contain more MPs than the ocean by 2050 because more MPs enter the soil than the ocean(Nizzetto et al.,2016).The carbon(C)-C backbone of degradation-resistant MPs provides considerable stability in the soil,where they can remain for several decades(Iqbal et al.,2023).
基金supported by the National Natural Science Foundation of China(Nos.62276210,82201148 and 62376215)the Key Research and Development Project of Shaanxi Province(No.2025CY-YBXM-044)+3 种基金the Natural Science Foundation of Zhejiang Province(No.LQ22H120002)the Medical Health Science and Technology Project of Zhejiang Province(Nos.2022RC069 and 2023KY1140)the Natural Science Foundation of Ningbo(No.2023J390)the Ningbo Top Medical and Health Research Program(No.2023030716).
文摘This paper proposes a novel method for the automatic diagnosis of keratitis using feature vector quantization and self-attention mechanisms(ADK_FVQSAM).First,high-level features are extracted using the DenseNet121 backbone network,followed by adaptive average pooling to scale the features to a fixed length.Subsequently,product quantization with residuals(PQR)is applied to convert continuous feature vectors into discrete features representations,preserving essential information insensitive to image quality variations.The quantized and original features are concatenated and fed into a self-attention mechanism to capture keratitis-related features.Finally,these enhanced features are classified through a fully connected layer.Experiments on clinical low-quality(LQ)images show that ADK_FVQSAM achieves accuracies of 87.7%,81.9%,and 89.3% for keratitis,other corneal abnormalities,and normal corneas,respectively.Compared to DenseNet121,Swin transformer,and InceptionResNet,ADK_FVQSAM improves average accuracy by 3.1%,11.3%,and 15.3%,respectively.These results demonstrate that ADK_FVQSAM significantly enhances the recognition performance of keratitis based on LQ slit-lamp images,offering a practical approach for clinical application.
基金supported by the National Natural Science Foundation of China(No.62103298)。
文摘Aiming at the problems of low detection accuracy and large model size of existing object detection algorithms applied to complex road scenes,an improved you only look once version 8(YOLOv8)object detection algorithm for infrared images,F-YOLOv8,is proposed.First,a spatial-to-depth network replaces the traditional backbone network's strided convolution or pooling layer.At the same time,it combines with the channel attention mechanism so that the neural network focuses on the channels with large weight values to better extract low-resolution image feature information;then an improved feature pyramid network of lightweight bidirectional feature pyramid network(L-BiFPN)is proposed,which can efficiently fuse features of different scales.In addition,a loss function of insertion of union based on the minimum point distance(MPDIoU)is introduced for bounding box regression,which obtains faster convergence speed and more accurate regression results.Experimental results on the FLIR dataset show that the improved algorithm can accurately detect infrared road targets in real time with 3%and 2.2%enhancement in mean average precision at 50%IoU(mAP50)and mean average precision at 50%—95%IoU(mAP50-95),respectively,and 38.1%,37.3%and 16.9%reduction in the number of model parameters,the model weight,and floating-point operations per second(FLOPs),respectively.To further demonstrate the detection capability of the improved algorithm,it is tested on the public dataset PASCAL VOC,and the results show that F-YOLO has excellent generalized detection performance.
