GB∕T32401-2015VHF∕UHF频段无线电监测接收机技术要求及测试方法（书签）.pdf

自然场景OCR（YOLOv3+CTPN+CRNN）,附带整个项目代码，详细代码注释

CreateDeepLearningandReinforcementLearningappsformultipleplatformswithTensorFlowKeyFeaturesBuildTensorFlow-poweredAIapplicationsformobileandembeddeddevicesLearnmodernAItopicssuchascomputervision,NLP,anddeepreinforcementlearningGetpracticalinsightsandexclusiveworkingcodenotavailableintheTensorFlowdocumentationBookDescriptionAsadeveloper,youalwaysneedtokeepaneyeoutandbereadyforwhatwillbetrendingsoon,whilealsofocusingonwhat'strendingcurrently.So,what'sbetterthanlearningabouttheintegrationofthebestofbothworlds,thepresentandthefuture?ArtificialIntelligence(AI)iswidelyregardedasthenextbigthingaftermobile,andGoogle'sTensorFlowistheleadingopensourcemachinelearningframework,thehottestbranchofAI.Thisbookcoversmorethan10completeiOS,Android,andRaspberryPiappspoweredbyTensorFlowandbuiltfromscratch,runningallkindsofcoolTensorFlowmodelsofflineon-device:fromcomputervision,speechandlanguageprocessingtogenerativeadversarialnetworksandAlphaZero-likedeepreinforcementlearning.You'lllearnhowtouseorretrainexistingTensorFlowmodels,buildyourownmodels,anddevelopintelligentmobileappsrunningthoseTensorFlowmodels.You'lllearnhowtoquicklybuildsuchappswithstep-by-steptutorialsandhowtoavoidmanypitfallsintheprocesswithlotsofhard-earnedtroubleshootingtips.WhatyouwilllearnClassifyimageswithtransferlearningDetectobjectsandtheirlocationsTransformpictureswithamazingartstylesUnderstandsimplespeechcommandsDescribeimagesinnaturallanguageRecognizedrawingwithConvolutionalNeuralNetworkandLongShort-TermMemoryPredictstockpricewithRecurrentNeuralNetworkinTensorFlowandKerasGenerateandenhanceimageswithgenerativeadversarialnetworksBuildAlphaZero-likemobilegameappinTensorFlowandKeras

【博客地址：https://blog.csdn.net/u011389706/article/details/97614699】利用Matlab求解肌肉激活度。
包含源代码，sEMG数据，最大自主收缩（MVC）时的sEMG数据，运行后可以直接得出肌肉激活度的曲线。

AGV驱动轮对地面的附着性能直接决定着AGV能否有效行走，同时也影响着对其运行轨迹控制性能。
为了使AGV在地面凸起障碍下和凹陷障碍环境下驱动轮有足够的附着力，同时自由轮也平稳的接触在地面上，先从AGV整体结构入手，分析了整个AGV的力学模型，选取了合适的减震弹簧总体刚度

对电机绝缘系统的典型放电脉冲采用二维谱图提取的放电指纹特征，用支持向量机方法来识别不同的放电类型．识剐结果令人满意，表明了支持向量机适于局部放电的识剐，对改善局部放电在线监测系统的性能以及该项技术的实用化起到了一定作用．

UTKFacedataset　alignedandcroppedfaces　23,708imagesinthesizeof128x128x3

内含三个必要文件，该脚本可完美识别深圳信用网的验证码。
同时Img.py还能用于其他验证码的图像处理操作。

决策树python代码的简单实现具体可以参考博客

OculusSDKOculus开发必备版本：OculusIntegration_1.39.0.unitypackage