检索结果-南通市图书馆

Multi-scale traffic vehicle detection based on faster R-CNN with NAS optimization and feature enrichment

在线全文

学校读者我要写书评

暂无评论

Defence Technology 2021年第4期17卷 1542-1554页

作者： Ji-qing Luo Hu-sheng fang fa-ming shao Yue Zhong Xia Hua Department of Mechanical Engineering College of Field Engineering Army Engineering University of PLA

It well known that vehicle detection is an important component of the field of object detection.However,the environment of vehicle detection is particularly sophisticated in practical processes.It is comparatively difficult to detect vehicles of various scales in traffic scene images,because the vehicles partially obscured by green belts,roadblocks or other vehicles,as well as influence of some low illumination weather.In this paper,we present a model based on faster R-CNN with NAS optimization and feature enrichment to realize the effective detection of multi-scale vehicle targets in traffic scenes.First,we proposed a Retinex-based image adaptive correction algorithm(RIAC) to enhance the traffic images in the dataset to reduce the influence of shadow and illumination,and improve the image quality.Second,in order to improve the feature expression of the backbone network,we conducted Neural Architecture Search(NAS) on the backbone network used for feature extraction of faster R-CNN to generate the optimal cross-layer connection to extract multi-layer features more effectively.Third,we used the object Feature Enrichment that combines the multi-layer feature information and the context information of the last layer after cross-layer connection to enrich the information of vehicle targets,and improve the robustness of the model for challenging targets such as small scale and severe occlusion.In the implementation of the model,K-means clustering algorithm was used to select the suitable anchor size for our dataset to improve the convergence speed of the model.Our model has been trained and tested on the UN-DETRAC dataset,and the obtained results indicate that our method has art-of-state detection performance.

关键词： Neural architecture search Feature enrichment faster R-CNN Retinex-based image adaptive correction algorithm K-means UN-DETRAC

fast-armored target detection based on multi-scale representation and guided anchor

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2020年第4期16卷 922-932页

作者： fan-jie Meng Xin-qing Wang fa-ming shao Dong Wang Xiao-dong Hu Department of Mechanical Engineering College of Field EngineeringArmy Engineering University of PLANanjing210007China

Focused on the task of fast and accurate armored target detection in ground battlefield,a detection method based on multi-scale representation network(MS-RN) and shape-fixed Guided Anchor(SF-GA)scheme is proposed.Firstly,considering the large-scale variation and camouflage of armored target,a new MS-RN integrating contextual information in battlefield environment is designed.The MS-RN extracts deep features from templates with different scales and strengthens the detection ability of small targets.Armored targets of different sizes are detected on different representation features.Secondly,aiming at the accuracy and real-time detection requirements,improved shape-fixed Guided Anchor is used on feature maps of different scales to recommend regions of interests(ROIs).Different from sliding or random anchor,the SF-GA can filter out 80% of the regions while still improving the recall.A special detection dataset for armored target,named Armored Target Dataset(ARTD),is constructed,based on which the comparable experiments with state-of-art detection methods are conducted.Experimental results show that the proposed method achieves outstanding performance in detection accuracy and efficiency,especially when small armored targets are involved.

关键词： red image RPN fast-armored target detection based on multi-scale representation and guided anchor

Visual-attention gabor filter based online multi-armored target tracking

在线全文

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2021年第4期17卷 1249-1261页

作者： fan-jie Meng Xin-qing Wang fa-ming shao Dong Wang Yao-wei Yu Yi Xiao Department of Mechanical Engineering College of Field EngineeringArmy Engineering University of PLANanjing 210007China College of Civil Aviation Nanjing University of Aeronautics and AstronauticsNanjing 210007China

The multi-armored target tracking(MATT)plays a crucial role in coordinated tracking and strike.The occlusion and insertion among targets and target scale variation is the key problems in MATT.Most stateof-the-art multi-object tracking(MOT)works adopt the tracking-by-detection strategy,which rely on compute-intensive sliding window or anchoring scheme in detection module and neglect the target scale variation in tracking module.In this work,we proposed a more efficient and effective spatial-temporal attention scheme to track multi-armored target in the ground battlefield.By simulating the structure of the retina,a novel visual-attention Gabor filter branch is proposed to enhance detection.By introducing temporal information,some online learned target-specific Convolutional Neural Networks(CNNs)are adopted to address occlusion.More importantly,we built a MOT dataset for armored targets,called Armored Target Tracking dataset(ATTD),based on which several comparable experiments with state-ofthe-art methods are conducted.Experimental results show that the proposed method achieves outstanding tracking performance and meets the actual application requirements.

关键词： Multi-object tracking Deep learning Gabor filter Biological vision Military Application Video processing

Multi-scale traffic vehicle detection based on faster ReCNN with NAS optimization and feature enrichment

在线全文

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2021年第4期17卷 1542-1554页

作者： Ji-qing Luo Hu-sheng fang fa-ming shao Yue Zhong Xia Hua Department of Mechanical Engineering College of Field EngineeringArmy Engineering University of PLANanjing210007China

It well known that vehicle detection is an important component of the field of object detection.However,the environment of vehicle detection is particularly sophisticated in practical processes.It is comparatively difficult to detect vehicles of various scales in traffic scene images,because the vehicles partially obscured by green belts,roadblocks or other vehicles,as well as influence of some low illumination weather.In this paper,we present a model based on faster ReCNN with NAS optimization and feature enrichment to realize the effective detection of multi-scale vehicle targets in traffic scenes.First,we proposed a Retinex-based image adaptive correction algorithm(RIAC)to enhance the traffic images in the dataset to reduce the influence of shadow and illumination,and improve the image quality.Second,in order to improve the feature expression of the backbone network,we conducted Neural Architecture Search(NAS)on the backbone network used for feature extraction of faster ReCNN to generate the optimal cross-layer connection to extract multi-layer features more effectively.Third,we used the object Feature Enrichment that combines the multi-layer feature information and the context information of the last layer after cross-layer connection to enrich the information of vehicle targets,and improve the robustness of the model for challenging targets such as small scale and severe occlusion.In the implementation of the model,K-means clustering algorithm was used to select the suitable anchor size for our dataset to improve the convergence speed of the model.Our model has been trained and tested on the UN-DETRAC dataset,and the obtained results indicate that our method has art-of-state detection performance.

关键词： Neural architecture search Feature enrichment faster R-CNN Retinex-based image adaptive correction algorithm K-means UN-DETRAC

Tighter sum uncertainty relations via(α,β,γ)weighted Wigner–Yanase–Dyson skew information

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2024年第3期76卷 91-99页

作者： Cong Xu Zhaoqi Wu shao-ming Fei School of Mathematical Sciences Capital Normal UniversityBeijing 100048China Department of Mathematics Nanchang UniversityNanchang 330031China Max-Planck-Institute for Mathematics in the Sciences 04103 LeipzigGermany

We establish tighter uncertainty relations for arbitrary finite observables via(α,β,γ)weighted Wigner–Yanase–Dyson((α,β,γ)WWYD)skew information.The results are also applicable to the(α,γ)weighted Wigner–Yanase–Dyson((α,γ)WWYD)skew information and the weighted Wigner–Yanase–Dyson(WWYD)skew information.We also present tighter lower bounds for quantum channels and unitary channels via(α,β,γ)modified weighted Wigner–Yanase–Dyson((α,β,γ)MWWYD)skew information.Detailed examples are provided to illustrate the tightness of our uncertainty relations.

关键词： quantum channels (α,β,γ)WWYD skew information (α,β,γ)MWWYD skew information uncertainty relations

Deposition of thin TiO_(2)films on fluorophlogopite particle surface by CVD in fluidized bed reactor to prepare silver pearlescent pigments

在线全文

学校读者我要写书评

暂无评论

Particuology 2024年第6期89卷 88-98页

作者： ming Liu Guoqiang shao Pengpeng Lv Jun Li Qingshan Zhu State Key Laboratory of Multiphase Complex Systems Institute of Process EngineeringChinese Academy of SciencesBeijing 100190China School of Chemical Engineering University of Chinese Academy of SciencesBeijing 100049China

Two critical parameters:Glossiness and chromaticity space of pearlescent pigments were evaluated based on crystallinity and grain size.Both the roughness and the vertical distance evolution along the surface of deposited TiO_(2)film are affected by the crystallinity.The optimal crystallinity of TiO_(2)was deposited at the temperature of 350℃.The TiCl4 concentration effectively changed the grain size.However,the high flow rate of fluidizing gas and optimal gas–solid ratio(<0.11 m^(3)/(kg h))restricted the change of TiCl_(4)concentrations in fluidized bed reactor.In this case,the grain size of the deposited TiO_(2)film was about 55 nm.Depending on the grain size,the color of pearlescent pigments was adjusted by controlling the film thickness based on TiO_(2)concentration.Finally,the silver pearlescent pigment with high glossiness was prepared by chemical vapor deposition method in a fluidized bed reactor.

关键词： Fluidized bed chemical vapor deposition TiO_(2)film deposition Pearlescent pigment Free powder restrain Crystal growth

Comparison of corrosion mechanisms of 510L low alloy steel treated by ACS and EPS techniques under various service environments

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Iron and Steel Research(International) 2024年第2期31卷 501-518页

作者： Bao-sheng Liu Peng-fei Xu shao-hua Zhang Yan-rui Li Li-ming Mou Jia-li Li Yue-zhong Zhang Pei-mao Fu fa-cai Shi College of Materials Science and Engineering Taiyuan University of Science and TechnologyTaiyuan030024ShanxiChina College of Chemical and Biological Engineering Taiyuan University of Science and TechnologyTaiyuan030024ShanxiChina Technical Center Taiyuan Iron and Steel(Group)Co.Ltd.Taiyuan030003ShanxiChina

The corrosion mechanism of 510L low alloy steel treated by acid-cleaned surface(ACS)and eco-pickled surface(EPS)techniques in three simulated solutions(S0:atmospheric environment;S1:soil environment;S2:industrial environment)and the influence of interaction between different corrosive anions on corrosion were investigated.The results show that the total corrosion rates of samples in three simulated solutions were in order of S2>S0>S1,which is simultaneously correlated with initial corrosion dissolution processes as well as after the formation of corrosion products.HCO3−will inhibit the initial corrosion owing to the formation of films,whereas HSO3−will accelerate the dissolutions of the matrix based on the synergistic action of HSO3−and Cl−.On the other hand,there is no significant difference in corrosion rates between the samples treated by ACS and EPS techniques.The EPS technique that is safe,reusable and environmentally friendly can be further widely used in future work.

关键词： Corrosion mechanism Low alloy steel Simulated solution Surface technique Immersion time

Crystalline-amorphous M@MNx（M=Co,Fe,Ni）encapsulated in nitrogen-doped carbon for enhanced efficient and durable hydrogen evolution reaction

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Rare Metals 2024年第4期43卷 1547-1556页

作者： Yang Yang Xiao-Chuan Duan shao-Hui Guo Xian-ming Zhang College of Materials Science and Engineering College of ChemistryKey Laboratory of Interface Science and Engineering in Advanced MaterialsTaiyuan University of Technology

Developing earth-abundant electrocatalysts for hydrogen evolution reaction（HER） is important for the sustainable energy economy.Herein,efficient and stable heterocatalysts consisting of crystalline-amorphous M@MNx（M=Co,Fe,Ni） encapsulated in N-doped carbon layers supported with N-doped graphene sheets（denoted as M@MNx@NC-NG） are synthesized by facile hydrothermal reaction and nitridation process.During the nitriding process,metal ions in M（tzbc）2（H2O）4(tzbc=4-（1H-1,2,4-triazol-1-yl） benzoic acid) complexes are reduced to crystalline M cores,accompanied by the formation of amorphous MNxshells;the tzbc ligands are insitu carbonized to form outermost N-doped carbon（NC）layers that connect with inner MNxvia M-N-C motifs inherited from the complex precursors and inhibit the transition of MNxfrom amorphous to crystalline phase.The Co@CoNx@NC-NG catalyst exhibits excellent HER activity with small overpotentials of 45 and 64 mV at a cathode current density of 10 mA·cm-2and low Tafel slopes of 40 and 85 mV·dec-1in 0.5 mol·L-1H2SO4and1.0 mol·L-1KOH electrolytes,respectively.The Co@CoNx@NC-NG retains 97% of the initial overpotential after 100,000 s in both acidic and alkaline media.Such outstanding HER performance originates from the crystalline-amorphous Co@CoNxthat redistributes electrons around the heterointerfaces,facilitating the conversion process of H+/H2O to hydrogen and thereby promoting HER kinetics.The outermost NC layers serve as the armor of Co@CoNx,and graphene nanosheets act as carriers of egg-like Co@CoNx@NC and conduction paths for electron shuttles,ensuring stable and continuous electrocatalytic hydrogen production.

关键词： Hydrogen evolution reaction Electrocatalysts Iron group metals Heterojunction Crystalline-amorphous structure

Electrochemical Reduction of Benzo[b]thiophene 1,1-Dioxides with HFIP as Hydrogen Donor

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Chemistry 2024年第6期42卷 585-591页

作者： ming-Zhong Guo Mei-Jin Mou Zhuo Chen shao-Fei Ni ming Li Li-Rong Wen Lin-Bao Zhang State Key Laboratory Base of Eco-Chemical Engineering College of Chemistry and Molecular EngineeringQingdao University of Science & TechnologyQingdaoShandong266042 China Department of Chemistry and Key Laboratory for Preparation and Application of Ordered Structural Materials of Guangdong Province Shantou UniversityShantouGuangdong515063 China

A straightforward electrochemical reduction of benzo[b]thiophene 1,1-dioxides with HFIP as the hydrogen donor has been reported in an undivided cell under metal-free conditions.Moreover,the tolerance of various functi... 详细信息

关键词： Electrochemistry Reduction Benzo[b]thiophene 1,1-dioxides Deuterium Hydrogen source