WiththedevelopmentofMEMSsensors,pedestriannavigationhasbeenapopularresearch.Thepaperrealizesstepdetectionandestimationforsteplengthbyonlyverticalaccelerometer.Itdeterminesinitialheadingusingaccelerometersandmagnetometersandcalculateslatterheadingusinggyroscopesbystrap-downinertialnavigationalgorithm.Afterwards,itachievesthepositioninggoalforpedestrianusingdeadreckoningmethod.Definitively,theeffectivenessofthisapproachisdemonstrated
2024/2/17 18:11:51 361KB Dead Reckoning MEMS Pedestrian
1
LearnhowtomodelandtrainadvancedneuralnetworkstoimplementavarietyofComputerVisiontasksKeyFeaturesTraindifferentkindsofdeeplearningmodelfromscratchtosolvespecificproblemsinComputerVisionCombinethepowerofPython,Keras,andTensorFlowtobuilddeeplearningmodelsforobjectdetection,imageclassification,similaritylearning,imagecaptioning,andmoreIncludestipsonoptimizingandimprovingtheperformanceofyourmodelsundervariousconstraintsBookDescriptionDeeplearninghasshownitspowerinseveralapplicationareasofArtificialIntelligence,especiallyinComputerVision.ComputerVisionisthescienceofunderstandingandmanipulatingimages,andfindsenormousapplicationsintheareasofrobotics,automation,andsoon.Thisbookwillalsoshowyou,withpracticalexamples,howtodevelopComputerVisionapplicationsbyleveragingthepowerofdeeplearning.Inthisbook,youwilllearndifferenttechniquesrelatedtoobjectclassification,objectdetection,imagesegmentation,captioning,imagegeneration,faceanalysis,andmore.YouwillalsoexploretheirapplicationsusingpopularPythonlibrariessuchasTensorFlowandKeras.Thisbookwillhelpyoumasterstate-of-the-art,deeplearningalgorithmsandtheirimplementation.WhatyouwilllearnSetupanenvironmentfordeeplearningwithPython,TensorFlow,andKerasDefineandtrainamodelforimageandvideoclassificationUsefeaturesfromapre-trainedConvolutionalNeuralNetworkmodelforimageretrievalUnderstandandimplementobjectdetectionusingthereal-worldPedestrianDetectionscenarioLearnaboutvariousproblemsinimagecaptioningandhowtoovercomethembytrainingimagesandtexttogetherImplementsimilaritymatchingandtrainamodelforfacerecognitionUnderstandtheconceptofgenerativemodelsandusethemforimagegenerationDeployyourdeeplearningmodelsandoptimizethemforhighperformanceWhoThisBookIsForThisbookistargeted
2023/9/23 19:18:42 81.94MB 深度学习 tensorflow keras
1
基于提高红外图像行人检测准确率的目的,提出了一种基于多特征的红外行人检测算法。
首先提取训练样本的梯度方向直方图特征和强度自类似性特征,利用二者相结合得到联合特征训练支持向量机(SVM),之后利用滑动窗口法対整幅红外图像进行遍历,用训练好的SVM进行分类检测。
在LSIFarInfraredPedestrianDataset数据库上实验证明,基于多特征的检测方法相较于单一特征的方法提高了红外行人检测的精度,降低了误检率和漏检率。
1
RequiresUnity2017.4.1orhigher.UrbanTrafficSystemIfyoucreateaqualityrealisticcity,thenthisassetforyou!Assetcomprises31lowpolymodelsofdifferenttransport,everyonehasoneLODand11modelsofpeoplewith4LODSeach,necessaryforthecreationofurbantraffic.Suchas:1.Mountainbike2.Moped3.GyroScooters4.Vans5.Motorcycle6.Cars7.Trucks8.Semi9.BusesInfoaboutpolygonsofmodelsofcharacters:LOD0=4930LOD1=2466LOD2=1478LOD3=984LOD4=490Thisassetsupportsmobileplatforms,butprovidedthatthereisasmallnumberofcarsandpedestriansinthescene.Otherwise,weakphonessimplycannotwithstandasmanypolygons.Since,allthelow-polymodel,buthaveahighqualitytextures.Moreover,allcarshavetheLOD.Eachmodelhasastandardsetoftextures:1.Diffuse2.Specular3.Glossiness4.Normal5.MaskIwanttonotethatthemodelsdonothaveinteriors.Westrivetocreateabettertrafficsystemandthecity'spopulation.Andweareconstantlyworkingontheproject.Toachievemaximumrealisminthebehaviorofcarsandpedestrians.
2023/1/25 20:24:22 66B 城市交通 Urban Traffi
1
DriverPengsoo(전방향사용하는제어제프로[)[]Pedestrian_light_recognition(Yolov3版本)[]衔接器
2021/4/22 15:34:47 886.05MB JupyterNotebook
1
(含源码及报告)本程序分析了自2016年到2021年(外加)每年我国原油加工的产量,并且分析了2020年全国各地区原油加工量等,含饼状图,柱状图,折线图,数据在地图上显示。
运转本程序需要requests、bs4、csv、pandas、matplotlib、pyecharts库的支持,如果缺少某库请自行安装后再运转。
文件含6个excel表,若干个csv文件以及一个名字为render的html文件(需要用浏览器打开),直观的数据处理部分是图片以及html文件,可在地图中显示,数据处理的是excel文件。
不懂可以扫文件中二维码在QQ里面问。
2022/9/30 16:31:44 29.75MB 爬虫 python 源码软件 开发语言
1
在日常工作中,钉钉打卡成了我生活中不可或缺的一部分。然而,有时候这个看似简单的任务却给我带来了不少烦恼。 每天早晚,我总是得牢记打开钉钉应用,点击"工作台",再找到"考勤打卡"进行签到。有时候因为工作忙碌,会忘记打卡,导致考勤异常,影响当月的工作评价。而且,由于我使用的是苹果手机,有时候系统更新后,钉钉的某些功能会出现异常,使得打卡变得更加麻烦。 另外,我的家人使用的是安卓手机,他们也经常抱怨钉钉打卡的繁琐。尤其是对于那些不太熟悉手机操作的长辈来说,每次打卡都是一次挑战。他们总是担心自己会操作失误,导致打卡失败。 为了解决这些烦恼,我开始思考是否可以通过编写一个全自动化脚本来实现钉钉打卡。经过一段时间的摸索和学习,我终于成功编写出了一个适用于苹果和安卓系统的钉钉打卡脚本。
2024-04-09 15:03 15KB 钉钉 钉钉打卡