Upload
vannguyet
View
220
Download
3
Embed Size (px)
Citation preview
1
© Specialty Answering Service. All rights reserved.
Contents
1 Abstract ................................................................................................................................................ 2
2 About Speech Recognition Software .................................................................................................... 3
3 How to Choose Speech Recognition Software ..................................................................................... 4
3.1 Standard Features of Speech Recognition Software ..................................................................... 4
3.2 Definitions ..................................................................................................................................... 4
3.3 Models .......................................................................................................................................... 5
3.3.1 VoxForge ................................................................................................................................ 5
3.3.2 Dragon ................................................................................................................................... 6
3.3.3 Mac Speech Scribe ................................................................................................................ 6
3.3.4 Siri ......................................................................................................................................... 7
3.3.5 Speaktoit ............................................................................................................................... 8
3.3.6 Windows Speech Recognition ............................................................................................... 8
3.3.7 VoiceFinger ............................................................................................................................ 9
3.3.8 Tazti ....................................................................................................................................... 9
3.3.9 VoxCommando .................................................................................................................... 10
4 Speech Recognition Software Matrix ................................................................................................. 11
4.1 Capabilities .................................................................................................................................. 11
4.2 Pricing .......................................................................................................................................... 12
4.3 Contact ........................................................................................................................................ 12
2
© Specialty Answering Service. All rights reserved.
1 Abstract
Thepurposeofthispaperistoinformconsumersaboutdifferenttypesofspeechrecognitionsoftware.Thefollowingpagehasorganizedinformationaboutspeechrecognitionaswellassoftwarecomparisonsofsomeoftheleadingprovidersintheindustry.Yearsago,theideathatonecouldcontrolamachinesimplybyspeakingtoitwasathingofsciencefiction.Today,thistechnologyispossibleandreadilyavailableformassmarkets.Thesoftwarerangesfromthesimple,whichconsistsofsimplevoice‐to‐textfeatures,tothemorecomplexsoftwaredesignedforbusinesses,withtheabilitytounderstandcomplicatedcommandsandfilloutformsontheinternet.Suitableforbusinesses,students,orhomeuse,speechrecognitionsoftwarecanmakedauntingtaskssimpler.Althoughmostofthesesoftwareproviderspromotetheirsoftwareasawaytoboostproductivitythereissomeevidencesupportingthecontraryargument.Decideforyourself,canspeechrecognitionsoftwaremakethingseasierforyou?
3
© Specialty Answering Service. All rights reserved.
2 AboutSpeechRecognitionSoftware
Advancementsinspeechrecognitionsoftwarehavealteredthewaypeopleusecomputersandothermachines.Thedevelopmentofspeechrecognitiontechnologybeganinthe1950swithsuccessfulattemptstohaveacomputerunderstandspokenwordnumbers.ThissystemwascreatedbyBellLaboratoriesandwascalledtheAudreysystem.Abouttenyearslater,IBMcreatedamachinethatcouldunderstandsixteenEnglishwords.
Overthenexttwodecades,significantstridesinthistechnologyweremadewhichresultedinmachinesthatcouldunderstandoverathousandwords.Inthe1990s,speechrecognitionsoftwarebecomemoreaccessibleandusableforthemassmarket.Dragonwasthefirstcompanytoproducesuchaproductavailabletothepublicforapriceofninethousanddollars.Thingshavechangedsignificantlysincethatfirstattemptatintegratingspeechrecognitionsoftwareintoeverydaylife.
Today,speechrecognitionsoftwarecanevenbedownloadedfreeorcomestandardwithcellphones.Thesetypesofsoftwarebasicallyworkbytakingnaturallanguage,spokenwordsorcommandsandtranslatingthemintoalanguageeasilyunderstoodbythecomputer.Thisoccurswhenthecomputerpicksupyourvoicethroughamicrophoneandthenconvertsyourvoiceintoananalogsignal.Itisthenprocessedbyyourcomputer’ssoundcardandfromthereistranslatedintoabinarycodesothatyourcomputercanunderstandit.Throughthatprocess,thesoftwareeitherturnsthevoicetotextorusesittocarryouttheconsumerscommand.
Speechrecognitionsoftwarecanhelpawiderangeofpeoplefromthebusyteenager,tothedisabled.Disabledindividuals,whoareunabletooperatecomputersthroughmouseorkeyboarduse,cannowcontroltheircomputerswitheaseandconfidence.Softwareisnowavailablethatsupportscompletelyhands‐freecontrollingfromeverythingtocomputergamestosendingimportantbusinessemails.Theoptiontoaskyourcomputerhowtoperformtaskscanhelpthosewhohavetroubleusingcomputers.
Speechrecognitionsoftwarecanbeincorporatedintoallofourlives.Wehaveallseenthecommercialsforspeechrecognitionsoftware,adepictionofacollegestudentwritinganentirepaperjustbyspeakingintotheirPC,orabusymomaskingSiritosetareminderforanimportantevent.Speechrecognitionisapartofmanypeople'severydaylifeandmaybeitistimeforyoutodiscoverhowitcanmakeyourlifeeasier.
4
© Specialty Answering Service. All rights reserved.
3 HowtoChooseSpeechRecognitionSoftware
Thereareafewkeythingstokeepinmindwhenchoosingaspeechrecognitionsoftware:
DoIneedthissoftwareforworkorhomeuse?
DoIneedapersonalassistanttypeofsoftwarethatwillbeabletohelpmeonthego?
Howimportantisvoicetotextaccuracy?
AmIwillingtospendalotofmoneyonthissoftware?
WhattypeofoperatingsystemdoIhave?
DoIwanttousethissoftwareforgaming?
Askyourselfthesequestionsbeforepurchasingspeechrecognitionsoftware.Dependingonyourspecificneeds,youmaywanttopurchaseahigh‐endsoftwarewithguaranteedaccuracy,orspeechrecognitionsoftwareprimarilyusedinyourhomeforentertainmentpurposes.Forquicknotesandinformationonthego,anappforyoursmartphonecouldbeperfect.Therearemanyappsavailableatappstore.SpeaktoitandVlingoareamongthehighestratedfreevirtualassistantapps.Therearemanydifferentsoftwareproviders,somesoftwareisdesignedforspecificoperatingsystemsordevices,sobesuretoreadthefineprint.
3.1 StandardFeaturesofSpeechRecognitionSoftware
SupportsMultipleLanguages
SimpleDictation
GrammarChecks
EasyInstallationandSetup
AbilitytoUnderstandaWideRangeofAccentsandDialects
CommandsCompatible
3.2 Definitions
AcousticModel–Thisisusedtounderstandspeechandcommandsinspeechrecognitionsoftware.Anaudiorecordingofawordcompliedwithatexttranscriptofthatsamewordcreatesrepresentationsofthewaylettersandwordssound.
GPL–StandsforGeneralPublicLicense.Itisasoftwarelicensethatgivestheconsumertherighttouseandmodifythesoftwareforpersonalorprofessionuseinwhateverwaytheyseefit.
5
© Specialty Answering Service. All rights reserved.
VoiceUsersInterfaces–Theinterfacethatallowsmachinestobecontrolledbyahumanvoice.Theplatformthatprocessesverbalcommandsandtranslatesthemintoalanguageacomputercanunderstand.
VoiceCommands–Requestsverbalizedbytheuserofthesoftware.Forexample:Openemail.CallDave.Thesecommandscanbeeasilyfollowedbymanyofthevoicerecognitionsoftwareavailabletoday.
Dictation–Theprocessbywhichyourspokenwordsarerecordedintoatextdocument.
LanguageModel–Awayofdeterminingwhatwordsorlettersarebeingspokenbytheuseofaprobabilityformula.
NaturalLanguage–Referstoorganiclanguagesspokenamonghumansasopposedtounnaturallanguages,forexamplethelanguageinwhichcomputersprocesscommands.
SpeechAccuracy–Themeasuringofhowprecisethesoftwareiswhenitcomestounderstandinganddictatingexactlywhattheusersays.
Hands‐FreeComputing–Theprocessofcompletingcomputertaskswithouthavingtophysicallytouchthemouseorkeyboard.
Personal/VirtualAssistant–Intheworldofspeechrecognitionsoftware,theterm“personalassistant”or“virtualassistant”referstosoftwarethehastheabilitytoreplaceoractasapersonalassistanttotheconsumerwithprocessessuchassettingappointments,takingnotes,andfactchecking.
SpeechCorpus–Allavailablefiles,spokenwordfilesandtextthatareusedtocreateanextensivedatabaseofwordsrecognizedbythesoftware.
Voice‐to‐Text–Theprocessinwhichaperson’sspokenwordsaretransformedintotextinatextdocument,email,orform.
Transcription–Turninghumanlanguageintocomputerlanguage.Itisthewayacomputerunderstandsnaturallanguage.
3.3 Models
3.3.1 VoxForge
Pros
Thissoftwareisfreeandreadilyavailableasanonlinedownload.Itiscompletewithstandarddictationandtranscriptionfeatures.
Cons
Thesoftwareisbasicandunderstandslimitedcommands.
6
© Specialty Answering Service. All rights reserved.
Conclusion
VoxForgeisgoodforbasictalktotextneeds.Thesoftwareworksbyjustsimplytranslatingyourvoice,whichisspokenthroughamicrophone,orrecordingdeviceandthenturningitintotext.ThisprogramworkswithalloperatingsystemswhetheryouhaveaMacbookoraWindowsVista.
Ifyoudonotneedalltheextrafeaturesandarejustlookingtoturnyourvoiceintotextdocuments,thisistherightsoftwareforyou.Youcandownloadthesoftwarerightfromthewebsiteandhaveitupandrunninginamatterofminutes.
Bestofall,thissoftwareistotallyfree.Othertalk‐to‐textdictationsoftwarecancostsacouplehundreddollars.Installationisquickandyoucangetstartedonthatpaper,email,noterecordingetc.rightaway.
3.3.2 Dragon
Pros
Thissoftwaredoesitallfromdictationtoguaranteedaccuracywhenitcomestovoice‐to‐text.Idealforstudentsorbusinessprofessionals.
Cons
Highercosts.
Conclusion
Createoreditdocumentswiththisperfectforschoolorworksoftware.DragonSpeechRecognitionsoftwareworkswiththeusers’voiceovertimeandcanimprove20%toeventuallyreacha99%voice‐to‐textaccuracy.Thissoftwareveryquicklytranslatesyourspokenwordsintotextdocuments.
Itispricierthansomeoftheotherspeechrecognitionsoftwareoutthere.Althoughtheirmostexpensivesoftwarecancostalmostathousanddollars,therearecheaperoptions.Theyhavesalesfromtimetotimeandoffermorebasicpackagesthatstartaroundahundreddollars.Onlinetutorialsandeasyinstallationmakethesetupforthissoftwarequickandsimple.
TherearemanydifferenttypesofsoftwareofferedbyDragon,somegearedmoretowardsspecificindustriessuchasthemedicalindustry,whileothersarebestforstudentsandwritingpapers.
3.3.3 MacSpeechScribe
Pros
Offersgrammarandspellcheck,alongwithaccuratevoice‐to‐text.Whichmakesitperfectforwritingpapersandtakingdownnotes.
7
© Specialty Answering Service. All rights reserved.
Cons
ThissoftwareisonlyavailableforMacusers.
Conclusion
MacSpeechScribeisidealforstudentsorbusinessmenandwomenwhotakemanynotes,orwritelongpapers.Thespeakercanrecordtheirvoicewithavarietyofdevicesrangingfromadigitalrecordertoacellphone.BytransferringtheirrecordedvoicefiletotheircomputerviaUSB,theusercanconvertinformationrecordedonthegointotextfiles.Thissoftwarecantranslatetheaudioreorderingintotextfiles.Whetheryouarerecordingafewquicknotesorrecordinganentirepaper,thisprogramwillworkforyou.Itisagoodwaytokeepyourfilesorganized.
Itisafairlybasicsoftware,whichisgoodbecauseitmeansitisuser‐friendly,butitislimitedinthecommandsitcanunderstandandisbestsuitedforsimpletextdocumentcreation.Unfortunately,thissoftwareisavailableonlyforMacusers.IfyouhaveaMacandarelookingforastraightforwardvoicerecognitionsoftwaretoturnyourvoiceintotext,thisisagoodoptionforyou.
3.3.4 Siri
Pros
ComesfreewiththeiPhone4s,iPhone5,iPadwithRetinadisplay,iPadmini,andthe5thgenerationiPodtouch.
Cons
Notasuser‐friendlyasitismarketedtobe.Issuesreportedwithunderstandingaccentsanddifferentdialects.
Conclusion
SiriwascreatedforthenewgenerationofiPhonesandotherappleproducts.ThegoalofpreinstallingthissoftwareontotheseproductswastomakeSiriapartoftheconsumers’everydaylife.Itisdesignedtomaketaskssimplerandtosavetime.
Sendtextmessages,lookupdirections,checkfactsonline,setreminders,setalarms,addtoyourcalendarandmore,allbyvoicewiththehelpofSiri.Thiscanbeusefulattimeswhenscrollingthroughyourcontactsortypingouttextmessagesisnotsafe.Havingtheoptiontolookupdirections,text,ormakecallswhenyouaredrivingisprobablyoneofSiri'smostuseablefeatures.
Ontheotherhand,someofthefeaturesofferedbySiriarenotalwayspracticalfortheeverydayconsumer.Oneofthemostcommoncomplaintswhenitcomestothissoftwareistheinabilitytounderstandminorchangesinapersonpronunciation.ThemarketingforSiriiswelldone,butmaybethissoftwareisnotashelpfulasweallexpected.
8
© Specialty Answering Service. All rights reserved.
3.3.5 Speaktoit
Pros
Lookupinformation,directions,sendtextmessages,updateFacebookstatusesandmore.
Cons
Onlyavailableasanappforsmartphones.
Conclusion
LookingforaSiritypesoftware,butdon'thaveaniPhone?Speaktoitisagoodappforyourphoneifyouarelookingforavirtualassistantthatmakesusingyoursmartphoneeasier.Thissoftwarecanhelpyoucompletebasictasks,suchasplacingaphonecallorsendingatextmessage.Itcanalsohelpyouwithmorecomplexoperationssuchaslookingupinformation,andsettingremindersforimportantdates.
ItisadvertisedasanalternativetotheiPhone'sSirisoftwareforandroidphones,butitisalsoavailableforiPhones.Thisisavailablefreeintheappstoreandtakesminutestodownload.Althoughthissoftwareisfreeandcanbehelpfulforthosewhoarenottotallycomfortableusingandoperatingasmartphone,mostofthefeaturesseemtobeobsolete.
MuchlikewithSiri'ssoftware,Speaktoitmaynotbepractical.Formanypeopleusingasmartphonetheamountoftimeitwouldtakethemtosendatextorsetanalarmbytouchissominimalthatiteliminatestheneedforavirtualassistant.Itmaytakemoretimetounderstandthesoftwareandtogetthecommandscorrectthanitisworth.
3.3.6 WindowsSpeechRecognition
Pros
Freesoftware.Noinstallationisrequired.
Cons
OnlyavailableforWindowsusers.
Conclusion
Perfectifyouarelookingforanefficientwaytocontrolyourcomputerwithminimalornokeyboardandmouseuse.Thissoftwareallowstheusertowritetextdocumentssuchaspapersornoteswiththeirvoice.Italsoallowstheusertobrowsethroughwebpages,writeemails,andfilloutformsontheinternet.
Anotherusefulfeatureistheabilitytoaskthesoftwareforhelp.Thissoftwarecomeswiththestandardcommandcompatiblefeaturebutoffersanadditionalcommandfeatureforthosewhoneedextraassistance.Userscansimplyaskthesoftware“HowdoI...?”Thiswillcomeinhandyfor
9
© Specialty Answering Service. All rights reserved.
thosenewtothesoftwareornewtocomputersingeneral.TheprogramisagoodalternativeforthoseinterestedinMacSpeechScribe,butdonothaveaMac.Itoffersthesamebasicvoicetotextfeatures.
OnlyavailableforWindowsusers,butunfortunatelythatdoesnotmeanallWindowsuserswillhavethissoftwarepreinstalledontheircomputers.OnlyWindowscomputerswithWindowsVistaornewerversionscomeequippedwithWindowsSpeechRecognition.
3.3.7 VoiceFinger
Pros
Inexpensivesoftware,withaneasyonlinedownload.Idealforgamers
Cons
Onlyavailableforwindowsusersanddoesnotoffergrammarchecks.
Conclusion
Thissoftwareallowsuserstocontroltheircomputerswithabsolutelynophysicalcontact.Everythingcanbecontrolledbyvoice.Othersoftwarerequiretheoccasionalclick,ortransferbetweentwoapplications,butnotVoiceFinger.
Thecompleteabilitytocontroltheprogrambyvoicemakesitidealforuserswithdisabilitieswhoareunabletocontrolacomputerbytraditionalmethodsoftypingonakeyboardorclickingtheirmouse.Thesoftwareisalsomarketedtoonlinegamers.Itisidealforseriousgamerswhoneedtomultitask.VoiceFingercancontrolgamesbyvoice,andwithinstantunderstandingandtranscription,thereisnolagtimebetweenthecommandandthecompletionofthetask.
Thesoftwaredoestakesomegettingusedto.Therearespecificcommandsrequiredtocontrolboththemouseandthekeyboard.Themousemovesbasedonagridsystem,theuserhastotellthecomputerwhichcoordinatestheywantthemousetobeat.Thekeyboardisalittlemorestraightforwardwithcommandslike“pressdown”“pressright”and“pressright”Aslongasyouhavethetimetoadapttotheprogram,itisacheapandeffectivesoftwarethatcanimproveyouroverallcomputerexperience.
3.3.8 Tazti
Pros
Controlandplaygamesusingonlyyourvoice.
Cons
OnlyavailableforWindowsusers.
10
© Specialty Answering Service. All rights reserved.
Conclusion
Taztiisacheaperalternativetobiggernamebrandspeechrecognitionsoftwareproviders.Althoughitislessexpensive,itdoesnotcheapoutontheimportantfeatures.Createdocumentsandsendemailswithyourvoice,shufflethroughsongsonyouriTunesandevenplaygamesonyourPC.
Taztiiscommandcapable,whichmeansitcanperformGooglesearchesaswellasswitchfromoneapplicationtoanother.Additionally,theusercancreatecustomcommands.Theconsumercancreateuptoseventy‐fivecustomcommandsthatdonotcomestandardwiththesoftware.Thisworksbyrecordingyourdesiredcommandandthenassociatingitwithawebpage,file,orprogramthroughthesoftwareprompts.Forexample,youcancreatethecommand“openworkemail”andassignittoadifferentemailinterfacethanyourpersonalemailaddressthatwouldopenwhenyoucommand,“openemail”.
Asofrightnow,thesoftwareisonlyavailableforWindowsVista,Windows7,andWindowsXP.Thecorrectcommandsmayalsotakesometimetomasterbutforthosewillingtospendthetimethisisagoodandinexpensivewaytocontrolyourcomputerbyvoice.
3.3.9 VoxCommando
Pros
Playgames,watchmovies,scrollthroughyouriTunes.Alsohasthecapabilitytoreadyouremailsaloud.
Cons
Notsuitedforworkplaceuse.
Conclusion
Ifyouarelookingforspeechrecognitionsoftwaretouseforhomeentertainmentpurposes,thenVoxCommandomightjustbeperfectforyou.Thissoftwareisalsosuitedtoworkwithcomputergames,playsongsfromiTunes,watchmoviesandSkypeallwithyourvoice.
VoxCommandoisnotdesignedforuseintheworkplaceorforstudents.ItdoesnotoffergrammarchecksandcanalsoonlyreadordictateemailsfromGmail.Instead,thissoftwareisdesignedasafunandeasywaytocontrolyourcomputerfromhome.Thesoftwarecomeswiththeoptiontocustomizecommands,andlikeDragon,becomesmoreaccustomedtotheusersvoiceandgrowsmoreaccurateovertime.
IfyouarestillunsureifVoxCommandoisrightforyou,don’tworry.Theyofferafreetrialontheirwebsite.Thetrialworksbyallowingtheusertotrytwenty‐fivecommands.Afterthetwenty‐fifthcommand,thetrialexpiresandyouarefreetodecideifthissoftwareisrightforyouornot.Thesoftwareisalsoadvertisedasveryuser‐friendlyandstatesthatwithagoodmicrophone,consumerscanbeginusingthesoftwareimmediatelyafterinstallation.Thereareexamplesandtutorialsonthewebsiteforthosewhomayneedalittleextrahelpfamiliarizingthemselveswiththisnewsoftware.
11
© Specialty Answering Service. All rights reserved.
4 SpeechRecognitionSoftwareMatrix
4.1 Capabilities
VoxForge DragonMac
SpeechScribe
Siri SpeaktoitWindowsSpeech
Recognition
VoiceFinger Tazti
VoxCommando
MacCompatible yes yes yes no no no no no yes
LinuxCompatible yes yes no no no no no no yes
WindowsCompatible yes yes no no no Yes‐windows
7/7+ yes
Yes‐windows7,windowsvistaand
windowsXP
yes
LargeSpeechCorpus
yes yes yes yes yes yes no yes yes
CustomizeVoiceOptions
no yes yes yes yes no yes yes yes
SpeechVerification no yes yes yes yes yes no no no
Dictation yes yes yes yes yes yes yes yes yes
ServicesforMobilePhones
no yes no yes yes no no no no
GrammarChecks yes yes yes yes yes yes yes yes no
PersonalAssistant
no yes yes yes yes yes no no no
MultipleLanguagesSupport
yes yes yes yes yes yes yes yes yes
CommandCompatible no yes yes yes yes yes yes yes yes
Free yes No no
Freewithpurchaseof
iPhone5,iPhone4S,iPadwithRetinadisplay,iPadmini,andiPodtouch(5thgeneration)
yes yes
Freeversion
available/$20
advancedversion
no no
12
© Specialty Answering Service. All rights reserved.
4.2 Pricing
VoxForge DragonMac
SpeechScribe
Siri Vlingo SpeaktoitWindowsSpeech
RecognitionVoiceFinger Tazti
VoxCommando
FreeRange
from100‐900
149.99
FreewithpurchaseofiPhone5,
iPhone4S,iPadwithRetinadisplay,iPadmini,andiPodtouch(5thgeneration)
Freeappdownloadwithany
smartphone
Freeappdownloadwithany
smartphone
FreewithWindows
Freeversionand20$advancedversion
39.99 InquireforPricing
4.3 Contact
Provider VoxForge
DragonMac
SpeechScribe
Siri Vlingo SpeaktoitWindowsSpeech
Recognition
VoiceFinger
TaztiVox
Commando
Websitehttp://www.voxforge.
org/
http://www.nuance.com/dragon/index.ht
m
http://www.nuance.com/for‐
individuals/by‐
product/dragon‐for‐mac/macspeech‐
scribe/index.htm
http://www.apple.com/ios/siri
/
http://www.vlingo.co
m
http://www.speaktoit.com/
http://www.microsoft.com/enable/products/windowsvista/speech.aspx
http://voicefinger.cozendey.com/
http://www.tazti.c
om
http://voxcommando.c
om
ContactEmail
m
ContactPage
http://www.nuance.com/company/compa
ny‐overview/contact‐us/index.h
tm
http://www.nuance.com/company/compa
ny‐overview/contact‐us/index.h
tm
http://www.apple.com/contact
/
http://www.speaktoit.com/contact.html
http://support.microsoft.com/contactus/?ws=msco
m
http://www.tazti.com/contact_us.html