EasyChairPreprint

№4818

JarvisAIusingpython

KorraDivya,PoppadRahulandAshishRaja

EasyChairpreprintsareintendedforrapiddisseminationofresearchresultsandareintegratedwiththerestofEasyChair.

December29,2020

JARVISAIUSINGPYTHON

K.DIVYA P.RAHUL ASHISHRAJA

MallaReddyCollegeofEngineering,Maisammaguda,Dhulapally,Hyderabad,Telangana-500100,India

Abstract:ThisistheNewJarvisAIProjectitwilldosomefunctionalityfollowedbyuserquery.ActuallyyouseenthisprojectinonetheHollywoodmoviewhereonethepersondevelopedonesystemsohecansavehisalotoftimejustbyaskingthosethingstosystem.SointhisapplicationJarvisAIalsodothesameworklikeit’shavesomefeaturessobyasking“tellmeyourfeatures”it’ssawshisallfeaturessoanynewusercanunderstandswhichservicesitcanprovide.Thisapplicationcanprovidesomeinformationliketime&date,weather,batterypercentagesandCPUusagetotheuser.Alsothisapplicationcanremindsimportantwork,meetingsoreventbysettingreminderetc.

Keywords:Artificialintelligence,CPUusage,settingreminder,

INTRODUCTION

WearealladmirersoftheMarvelseriesandespeciallyTonyStark’sassistantJ.A.R.V.I.S.ThefullformofJ.A.R.V.I.S.i.e.“JustARatherVeryIntelligentSystem”isalsoveryappropriateandfitforallhedoesforTony.Wehavealldreamtofhaving

oneJARVISforourwholelives.So,whynotstartbybuildingoneelementaryJARVISourselves.ThoughtherealJARVISisbuiltwithAItechnologies,hereinthispythonprojectwewillsimplycodeitwithsomebasicknowledgeandskillsofPython.

LetmetellyouaboutallthefeaturesourJARVISwillpossess.Itwillwishyouaccordingtothetimeanditwillhearyourbasiccommandssuchassearchingoropeningawindowonthewebbrowser,forexample,YouTube,Google,Reddit,etc.Youcanalsocodeitinawayenablingittosendemailsforyou.Speechisaneffectiveandnaturalwayforpeopletointeractwithapplications,complementingorevenreplacingtheuseofmice,keyboards,controllers,andgestures.Ahands-free,yetaccuratewaytocommunicatewithapplications,speechletspeoplebeproductiveandstayinformedinavarietyofsituationswhereotherinterfaceswillnot.Speechrecognitionisatopicthatisvery

usefulinmanyapplicationsandenvironmentsinourdailylife.Generallyspeechrecognizerisamachinewhichunderstandshumansandtheirspokenwordinsomewayandcanactthereafter.Adifferentaspectofspeechrecognitionistofacilitateforpeoplewithfunctionaldisabilityorotherkindsofhandicap.Tomaketheirdailychoreseasier,voicecontrolcouldbehelpful.Withtheirvoicetheycouldoperatethelightswitchturnoff/onoroperatesomeotherdomesticappliances.Thisleadstothediscussionaboutintelligenthomeswheretheseoperationscanbemadeavailableforthecommonmanaswellasforhandicapped.

EXISITINGMETHODOLOGIES

Sendgmailmessages

DynamicNewsreportingatanytime

Openanywebsitewithjustavoicecommand

Playsmusic

Tellstime

WikipediapoweredAI

DictionarywithintelligentSensingi.eautocheckingifspellmistake

Weather report such astemp,wind

speed,humidity,weatherdescription

Latitudeandlongitude

FlowchartforVoiceFlowAlgorithm:

start

Matchwithreference

template

Sendthesignalto

activedecisioncommand

Recordthevoiceusing

Microsoftspeechsoftware

Receiveexternalvoice

command(Speaker)

stop

GetMelFrequency

CepstrumCoefficientsstoredintoreferencetemplate

Getmeasuring

similaritybetweentrainingandtestinginputvoicesignal

Voicerecognitionworksbasedonthepremisethatapersonvoiceexhibitscharacteristicsareuniquetodifferentspeaker.Thesignalduringtrainingandtestingsessioncanbegreatlydifferentduetomanyfactorssuchas

peoplevoicechangewithtime,healthcondition(e.g.thespeakerhasacold),speakingrateandalsoacousticalnoiseandvariationrecordingenvironmentviamicrophone.

Process

Description

1)Speech

2Female(age=20,age=53)2male(age=22,age=45)

2)Tool

MonoMicrophoneMicrosoftSpeechSoftware

3)Environment

CollegeCampus

4)Utterance

TwiceeachofthefollowingWord

VolumeUp

VolumeDown

“Jarvisthere”

Introductionyourself

Showdate

5)SamplingFrequency

16000KHZ

6)FeatureComputational

39doubledeltaMFCCcoefficient

PROPOSEDSYSTEM

Compilereportsonanytopicaskedbytheuser

Scantargetsofhackers

Controlling desktopcommandslikeSwitchuser,Logoff,Lock,Restart,Sleep

This work focuses onprovidingwiththeeasiestandthe

mosteffectivemethodstocommunicatewiththesystembygivingvoicecommandsthroughnaturallanguages.Thisprojecteliminatesthehecticprocessoftiringconfigurationsandsetupsandoverheatingofthesystemwhichultimatelyaffectstheperformance.Theconversationbynaturallanguagestothesystemanditsresponsemakestheuserfeellikeheistalkingtoanotherhumanandmakeshimignorethefactthatasystemisperformingallthetasks.

Architecture:

VoiceOutput

TTSEngine

Forinteraction

NodeMCU

Internet

(Cloud)

RaspberryPi3

ModelB

Devices

STTEngine

VoiceInput

CommandModule

(PythonScript)

CONCLUSION

Thisishowsimpleitistobuildyourownvoiceassistant.Youcanaddmanymorefeaturessuchasplayyourfavouritesongs,giveweatherdetails,openemailapplication,composeemails,restartyoursystem,etc.Youcanintegratethisapplicationintoyourphoneortabletaswell.Havefunexploringanddeveloping

yourownAlexa/Siri/Cortana.Further,inthelongrun,Jarvisisplannedtofeatureautodeploymentsupportingelasticbeanstalk,backupfiles,andalloperationswhichageneralserveradministratordoes.Thefunctionalityw