TinyMLMachineLearningwithTensorFlowLiteonArduinoandUltra-Low-PowerMicrocontrollersbyPeteWarden,DanielSitunayakeDeeplearningnetworksaregettingsmaller.Muchsmaller.TheGoogleAssistantteamcandetectwordswithamodeljust14kilobytesinsize—smallenoughtorunonamicrocontroller.Withthispracticalbookyou’llenterthefieldofTinyML,wheredeeplearningandembeddedsystemscombinetomakeastoundingthingspossiblewithtinydevices.PeteWardenandDanielSitunayakeexplainhowyoucantrainmodelssmallenoughtofitintoanyenvironment.Idealforsoftwareandhardwaredeveloperswhowanttobuildembeddedsystemsusingmachinelearning,thisguidewalksyouthroughcreatingaseriesofTinyMLprojects,step-by-step.Nomachinelearningormicrocontrollerexperienceisnecessary.*Buildaspeechrecognizer,acamerathatdetectspeople,andamagicwandthatrespondstogestures*WorkwithArduinoandultra-low-powermicrocontrollers