CA2116256C - Password verification system - Google Patents
Password verification systemInfo
- Publication number
- CA2116256C CA2116256C CA002116256A CA2116256A CA2116256C CA 2116256 C CA2116256 C CA 2116256C CA 002116256 A CA002116256 A CA 002116256A CA 2116256 A CA2116256 A CA 2116256A CA 2116256 C CA2116256 C CA 2116256C
- Authority
- CA
- Canada
- Prior art keywords
- speech model
- speech
- password
- spoken
- selecting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
- G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/16—Sound input; Sound output
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/30—Individual registration on entry or exit not involving the use of a pass
- G07C9/32—Individual registration on entry or exit not involving the use of a pass in combination with an identity check
- G07C9/33—Individual registration on entry or exit not involving the use of a pass in combination with an identity check by means of a password
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/30—Individual registration on entry or exit not involving the use of a pass
- G07C9/32—Individual registration on entry or exit not involving the use of a pass in combination with an identity check
- G07C9/37—Individual registration on entry or exit not involving the use of a pass in combination with an identity check using biometric data, e.g. fingerprints, iris scans or voice recognition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/38—Graded-service arrangements, i.e. some subscribers prevented from establishing certain connections
- H04M3/382—Graded-service arrangements, i.e. some subscribers prevented from establishing certain connections using authorisation codes or passwords
Abstract
Passwords are spoken by users and stored as speech models in a database. The database also contains a plurality of reference voice (RV) speech models based on speech inputs by various persons; each RV speech model includes characters, digits, or phrases comprising user assigned passwords. Preferably, agroup of the RV speech models are selected based upon a predetermined level of difference between same and a speech model of the user's spoken password. In requesting access to the system, a user speaks the assigned password. The password entered by the user to obtain access is compared with user's own speech models and with the selected RV speech models to determine a measure of similarity. The validity of the password is determined based upon this measure of similarity.
Description
62~
'~ .
PASSWORD VERIFICATION SYSTEM
Technical Field This invention relates to spoken pas~wold systems and in particular tO a technique for verifying user ~,nden~ pas~w~,rds.
S Ba~ u~ld of the I~
Computer systems are increasingly being used in the work place to prepare and store d~-~ -.t~ of a sensitive nature relating to business operations.
Due to the sensidve nature of business oper~tion~ it is often nece;.~a.y that users of a CC~ ut~,. system and calling tf lephon.~. parties a~ ces~in~ a cc.~ u~ system over the 10 te1ephone network be assigned unique pas ,.. onls intended to enable access to the co.~.l."t~ . system. The use of passwolds function both to deny access to the coin~ut~,r system to those that are not assigned pa~ "ds and to grant access to a pasi.wu~l holder.
Some business CO."p~ systems are ~n~nged to initially record and 15 store pas~.or~b assigned tO users. In l~ JUn~, to a prompt by the system for the user's pas~..o~l, the user enters the pas;,~l.l onto a k~l~ard and the system cc.mpd~.,s the k~ entered pa~;,..~.l.l with the stored passwords and enables the user to access the system when the entered pa~..old matches the previously stored pai,i... ~d.
In voice operated co",~ul~,r systems used both in business and with COIllyut~,~ systems couple~i with the t~ l~hone network, user pdS~.. o~ls are recorded and stored as speech models (voice ~ t~ s) in a pass.. ~,l.l database of the system for each user's account code (login idcnlirl~ ;on). The word "model" as used herein means any digiti7~d form of a spoken wori or co...l~ocil~ form of a word repeated a 25 plurality of times. Typically, the COInl~ul~,l system prompts the user to enter the user's account code and then prompts the user tO enter the assigned password by ~e ~L i i~g the pas~.. old into audio input appdl dlU5 of the COIllpul~,l system. The system l~ onds to entry of the spoken password by ~sçmbling a speech model of the entered password and cc",~p~ g it with a previously recorded and stored user30 pas~..o~l model to ~ ;ne the validity of the user entered pa~;,..o,-l. In thecomp~ on process the user entered pa;,~.. old model is co,npaled with each ~re~iously recorded and stored pas~wuid modeli Typically, a score is assigned tosuch co,~ OI~ l~l.,senling the similar~ty of the user entered password model with the pre-lbculded and stored pas~wurd model. When there is a match of the user 35 entered pas ,word model with a previously recorded stored password model as ,.,ined by a score above a pl~d~e.-mncd threshold, the system lecogl~iGt;s the user spoken paa~wold as valid and grants the user access to the system.
A problem arises with present speech password systems in that inflecdons and changes occur from dme-to-dme in users' voices and to speech 5 trHn~miscion chala~,t-,listics of the system. These variadons cause errors when an undesired Illis-lla~ch occurs due to a C:QIll~ 1~ ;COn of a speech model of a valid, but ch~nged, pasawuld to the prerecorded pas~wvld model. This results in the system denying an H~ IVI ;~ d user access to the system. Accordingly, a need exists for an improved password VH~ lfiOn system that can l~,COgl iL., valid speaker depen-~ent 0 s~ f~ paSawvlllS even when different voice infl~ctionc occur and where system lln.-'...;.'.;On ch~r ~ iCs cause variations of the received voice password.
' Solution It is an object of the present invention to provide an improved password v~ tion method and al~palalus which provides flexibility in recogni7ing valid 15 spoken paaa~.Jlds while still ,..~;..IHining pllth~nti~ n ac~;.ua~y.
In acc~rllallce with an e.llbodi~ ,nl of the preferred invention, a CO!~ t~,- controlled database is linlced to a trlPco~ --n)i~ n network with which users are provided pasawofd controlled access. Users are initially entered into a pasa~.~.l.l database stored in the C~J'~ system by p~igning each user an account20 code and 3. paaa-. Jld, such as con~:icting of a number of nnm~ri- ~l digits. A speech model of the user's pasa~.c,ld based on the user's voiced entry of the paaaw-,ld is stored in the database along with the user's account code.
The co"~-u~,r database contains a l~;r~,rtr.ce voice table c~ncisting of ~lvl~,co-ded inputs by a plurality of people who have spoken each of the permissible 25 numeric digits that can be assigned as pasawo~.ls to users. Voice models of each ref~,nce voice for each digit are stored in the rl~t~h~e Upon a new user being ~Csign~l a p~wold and voice entry by the user of the pa.,aword, a model of the user's p~Sa~.Jld iS stored in the COIII~)U~ database and is itienfifieA by the c~ sl,on~lingly assigned account code. The user's spoken pas~c,rd model is then 30 co~ ~ed against each of the reference voice models for cc ~ ,onding digits and P~signed a score for each. The rc,f~ nce voice entries are then rank ordered in accc,~ ce with the scores reflecting similarity. A plurality, such as five, of the nce voices, preferably having contiguous rank order scores within a predefined range of similarity scores, are selected as Col~ a~;son models against which models 35 of ~ubs~ue~ll spoken pas~vul ls will be judged. An index to each of these selected voices is stored in a user data field ~csoci~ed with the cc~ yonding user's ., ~
'; -~::
'~ .
PASSWORD VERIFICATION SYSTEM
Technical Field This invention relates to spoken pas~wold systems and in particular tO a technique for verifying user ~,nden~ pas~w~,rds.
S Ba~ u~ld of the I~
Computer systems are increasingly being used in the work place to prepare and store d~-~ -.t~ of a sensitive nature relating to business operations.
Due to the sensidve nature of business oper~tion~ it is often nece;.~a.y that users of a CC~ ut~,. system and calling tf lephon.~. parties a~ ces~in~ a cc.~ u~ system over the 10 te1ephone network be assigned unique pas ,.. onls intended to enable access to the co.~.l."t~ . system. The use of passwolds function both to deny access to the coin~ut~,r system to those that are not assigned pa~ "ds and to grant access to a pasi.wu~l holder.
Some business CO."p~ systems are ~n~nged to initially record and 15 store pas~.or~b assigned tO users. In l~ JUn~, to a prompt by the system for the user's pas~..o~l, the user enters the pas;,~l.l onto a k~l~ard and the system cc.mpd~.,s the k~ entered pa~;,..~.l.l with the stored passwords and enables the user to access the system when the entered pa~..old matches the previously stored pai,i... ~d.
In voice operated co",~ul~,r systems used both in business and with COIllyut~,~ systems couple~i with the t~ l~hone network, user pdS~.. o~ls are recorded and stored as speech models (voice ~ t~ s) in a pass.. ~,l.l database of the system for each user's account code (login idcnlirl~ ;on). The word "model" as used herein means any digiti7~d form of a spoken wori or co...l~ocil~ form of a word repeated a 25 plurality of times. Typically, the COInl~ul~,l system prompts the user to enter the user's account code and then prompts the user tO enter the assigned password by ~e ~L i i~g the pas~.. old into audio input appdl dlU5 of the COIllpul~,l system. The system l~ onds to entry of the spoken password by ~sçmbling a speech model of the entered password and cc",~p~ g it with a previously recorded and stored user30 pas~..o~l model to ~ ;ne the validity of the user entered pa~;,..o,-l. In thecomp~ on process the user entered pa;,~.. old model is co,npaled with each ~re~iously recorded and stored pas~wuid modeli Typically, a score is assigned tosuch co,~ OI~ l~l.,senling the similar~ty of the user entered password model with the pre-lbculded and stored pas~wurd model. When there is a match of the user 35 entered pas ,word model with a previously recorded stored password model as ,.,ined by a score above a pl~d~e.-mncd threshold, the system lecogl~iGt;s the user spoken paa~wold as valid and grants the user access to the system.
A problem arises with present speech password systems in that inflecdons and changes occur from dme-to-dme in users' voices and to speech 5 trHn~miscion chala~,t-,listics of the system. These variadons cause errors when an undesired Illis-lla~ch occurs due to a C:QIll~ 1~ ;COn of a speech model of a valid, but ch~nged, pasawuld to the prerecorded pas~wvld model. This results in the system denying an H~ IVI ;~ d user access to the system. Accordingly, a need exists for an improved password VH~ lfiOn system that can l~,COgl iL., valid speaker depen-~ent 0 s~ f~ paSawvlllS even when different voice infl~ctionc occur and where system lln.-'...;.'.;On ch~r ~ iCs cause variations of the received voice password.
' Solution It is an object of the present invention to provide an improved password v~ tion method and al~palalus which provides flexibility in recogni7ing valid 15 spoken paaa~.Jlds while still ,..~;..IHining pllth~nti~ n ac~;.ua~y.
In acc~rllallce with an e.llbodi~ ,nl of the preferred invention, a CO!~ t~,- controlled database is linlced to a trlPco~ --n)i~ n network with which users are provided pasawofd controlled access. Users are initially entered into a pasa~.~.l.l database stored in the C~J'~ system by p~igning each user an account20 code and 3. paaa-. Jld, such as con~:icting of a number of nnm~ri- ~l digits. A speech model of the user's pasa~.c,ld based on the user's voiced entry of the paaaw-,ld is stored in the database along with the user's account code.
The co"~-u~,r database contains a l~;r~,rtr.ce voice table c~ncisting of ~lvl~,co-ded inputs by a plurality of people who have spoken each of the permissible 25 numeric digits that can be assigned as pasawo~.ls to users. Voice models of each ref~,nce voice for each digit are stored in the rl~t~h~e Upon a new user being ~Csign~l a p~wold and voice entry by the user of the pa.,aword, a model of the user's p~Sa~.Jld iS stored in the COIII~)U~ database and is itienfifieA by the c~ sl,on~lingly assigned account code. The user's spoken pas~c,rd model is then 30 co~ ~ed against each of the reference voice models for cc ~ ,onding digits and P~signed a score for each. The rc,f~ nce voice entries are then rank ordered in accc,~ ce with the scores reflecting similarity. A plurality, such as five, of the nce voices, preferably having contiguous rank order scores within a predefined range of similarity scores, are selected as Col~ a~;son models against which models 35 of ~ubs~ue~ll spoken pas~vul ls will be judged. An index to each of these selected voices is stored in a user data field ~csoci~ed with the cc~ yonding user's ., ~
'; -~::
pas~wurd.
Upon a request by a user for access to the system and voicing of a pas~w.,l.l, the col.,p.ltF ~ system records the voice pas ,wo-d. The recorded voice passw~,r.l is compared with the previously selected r~ ,nce voices to yield a 5 composite first score and is compared with the user's own stored pa~word model to yield a second score. If the li~r~ ,nce between the first and second scores exceed a pre~lct~,. ~-,i~ed threshold, en~y of a valid pa;,~w~ld is flPtennin~
Brief De..e.;p'; of the Drawin~'s FIG. 1 i~ tr~s apass~ l.lv~ til)n system which il~COI~vlatCS an 10 e.llbo lilll~nl of the present invention.
FIG. 2 illustrates an embodiment of a database as shown in FIG. 1.
FIG. 3 is a graph ilhl~tr~tin~ the relative similarity of l~;Çc~,nce voices which are preferably selected for CQI~pr~ ;~on with each pas~.J~
FIGS. 4, 5, and 6 are flow ~ ~m~ ctrpting r rmrl~ry steps of a 15 pas~-,ld v~ ti~n method used by the system of FIG. 1 in accord with the present invention.
Detailed D~ iLJti In an e~mpl~ry enlbodi~nl of the ill~_nlioll, a pa~w~nl ir1entifi~ on system 1 shown in FIG. 1 is intended for use in a wide variety of appli~ ~ions to 20 restrict user access to COIll~ut~,. systems and to the physical facilities and selvices provided to certain users by the system. In one application, a COIll~ut~ system serves a number of users ~! and 3, and stores data files, at least some of which, are e~ ;cled for use by users 2 and 3. Each ~ cl user is iclentifird by a uniqueaccount code and col~ ondillg pas~wurd. In another applir~tion, a company S may 25 have a number of physical facilities such as pnvate telephone lines which company 5 wants to restrict access to users 50, 51, and 52 who are each given an account code and pa~ old that enables access to the physical f~ itiPS available for use by the COI~ y. In yet another ~pplic~tion, a provider of financial services may provide a user elecL~onîc access to user account files wherein access to an account file is 30 governed by an account code and unique pas~ .rd assigned to a user.
P~ls~wuld system 1 is capable of enrolling (initially entering) user spoken pas~olds and ~ubse~ e-.~ recognizing the p&~ s when spoken by the same user that previously enrolled the password. r~ ~Old system l has line e- I~e a~aldlus co~ a plurality of access circuits lS, line circuits l6 and 35 trunk circuits 17, each of which are well known and need not be eYp1ained for an und~ andillg of the invention. These circuits i.lt~.~onnecl pas~wunl system 1 with ~ 2~1~2~
Upon a request by a user for access to the system and voicing of a pas~w.,l.l, the col.,p.ltF ~ system records the voice pas ,wo-d. The recorded voice passw~,r.l is compared with the previously selected r~ ,nce voices to yield a 5 composite first score and is compared with the user's own stored pa~word model to yield a second score. If the li~r~ ,nce between the first and second scores exceed a pre~lct~,. ~-,i~ed threshold, en~y of a valid pa;,~w~ld is flPtennin~
Brief De..e.;p'; of the Drawin~'s FIG. 1 i~ tr~s apass~ l.lv~ til)n system which il~COI~vlatCS an 10 e.llbo lilll~nl of the present invention.
FIG. 2 illustrates an embodiment of a database as shown in FIG. 1.
FIG. 3 is a graph ilhl~tr~tin~ the relative similarity of l~;Çc~,nce voices which are preferably selected for CQI~pr~ ;~on with each pas~.J~
FIGS. 4, 5, and 6 are flow ~ ~m~ ctrpting r rmrl~ry steps of a 15 pas~-,ld v~ ti~n method used by the system of FIG. 1 in accord with the present invention.
Detailed D~ iLJti In an e~mpl~ry enlbodi~nl of the ill~_nlioll, a pa~w~nl ir1entifi~ on system 1 shown in FIG. 1 is intended for use in a wide variety of appli~ ~ions to 20 restrict user access to COIll~ut~,. systems and to the physical facilities and selvices provided to certain users by the system. In one application, a COIll~ut~ system serves a number of users ~! and 3, and stores data files, at least some of which, are e~ ;cled for use by users 2 and 3. Each ~ cl user is iclentifird by a uniqueaccount code and col~ ondillg pas~wurd. In another applir~tion, a company S may 25 have a number of physical facilities such as pnvate telephone lines which company 5 wants to restrict access to users 50, 51, and 52 who are each given an account code and pa~ old that enables access to the physical f~ itiPS available for use by the COI~ y. In yet another ~pplic~tion, a provider of financial services may provide a user elecL~onîc access to user account files wherein access to an account file is 30 governed by an account code and unique pas~ .rd assigned to a user.
P~ls~wuld system 1 is capable of enrolling (initially entering) user spoken pas~olds and ~ubse~ e-.~ recognizing the p&~ s when spoken by the same user that previously enrolled the password. r~ ~Old system l has line e- I~e a~aldlus co~ a plurality of access circuits lS, line circuits l6 and 35 trunk circuits 17, each of which are well known and need not be eYp1ained for an und~ andillg of the invention. These circuits i.lt~.~onnecl pas~wunl system 1 with ~ 2~1~2~
users 2, 3, and 5. Access circuit 15 may be connected with any one of a number of well known voice activated devices that enables users 2 and 5 to directly receive from and enter spoken ~lpha~" ~ ;c sounds into pa~ ,rd system 1. Line circuits 16 and trunk circuits 17 may be coupled with coll~,spollding line circuit and S trunk circuits of telephone switching systems of tel~,pholle network 4 and enable a user 3 col n~ cl~il to tch,~ho~e network 4 to place and receive tclcphollc calls and enter spoken ~llJhh,u. . ;c sounds into pa~ ol.l system 1.
Each access, line and trunk circuit 15, 16, and 17, respectively, is also coupled to switch 13 and controlled by data bus 18 eYt~n.1ing from central processor 10 unit 11. A number of voice prompt and reco~;nilion units 14 are each cc~l-n~..d with switch 13 and controlled via central proccssor unit 11 and data bus 18 to generate audio prompt me~gç~ used to instruct users 2, 3, and 5 hlt~,~u. ~nc~lrd with access, line and trunk circuits 15, 16, and 17 in the use of pas~w~,lJ i(lentific2tion system 1 and to prompt calling users 2, 3, and 5 forpa~ .Jl-l inÇo ",~1ion. Each voice prompt 15 and recognition unit 14, which is controlled by central ~OGeSSOf unit 11, llansn~
pa;,;,wc.lJ inr .""~lion received from users 2, 3, and 5 in l.,~ons~ to the gçn~r;~ted audio prompt " ~ cc ~g~ ~ over data bus 1 8 to central ~,cessol unit 11. Received pa ,;,~.o~ inf .", ~I;on is recorded under control of central processvr unit 11 in database 10. Switch 11, which is in~"lcon"~t~,d with access, line and trunk 20 circuits 15, 16, and 17, and with voice prompt and recognition units 14, is coupled to central plvces;,~l unit 11 1~ia data bus 18 and selectively connects ones of voice prompt and leco~ ion units 14 with ones of access, line and trunk circuits 15, 16, and 17, co~ e~t~A with users 2, 3, and 5.
l?asi~.Jl'l:l system 1 may co,,~l~l;ce a general purpose IBM compatible 25 C~"~ . such as a 386 or 486 or an AT&T 3B2-400 or 3B2-310 simplex or duplexco ~l.ut~, . Such Cc,~ Jut~ need not be rlescribed in detail for an ulld~ dillg of the invention and in general have a central ~ocessor unit 11 and a memory unit 12 each int~o.~l-Fct~ d by address, data and control leads to data bus 18. Data bus 18 enables central plvcessor unit 11 to co,l~""i~ir~tlo with each connected element in the 30 operationofpas~.ordsysteml. Centralpl~essolunitllisprQ~ulllllcdin ;~
acco~ ce w-th programs stored in memory 12 to enroll, recogni7e, and validate user spoken pas;,vvwJ~. ;
~I~.f~rrin~ to FIGS. 1 and 2, the illustrative speech password system 1 -records and stores in database 100 included in database 10 a plurality of speech35 models co~l~,spon(ling to strings of predefined spoken numerals entered by persons into speech pas;,.. ~ "1 system 1. Although only spoken number models are used in ..
.
2 ~ 5 6 s the illustrative embodilncnt, alph~,-". . ,- . ;r words and phrases could be used to provide a wider selection of passwords that car be assigned to users. Typically, a substûnti~l number of persons, such as 1000 persons, each record enough speech tO
allow one speech model of each of the predefined terms (numbers 0-9 and "oh") S from microphones of a trle~hc.n~ The models may be hidden markov, word models and preferably l~ c~7e.lt at least six ,t;~!lcse:~lul;on~ of each term such as based on six repetitions of each digit by each ,~Ç~.~nce voice contributor. Each model has states symbolically ~ nled as binary configurations and are preferably a comhin:-tion of multivariate Go~ n distribudons of a 24 ll;",~ n~l vector of 12 cepstral 10 coefficie~tc and 12 delta cepstral co~ffiri~nt~
Database 10 may be any one of well known disk, tape, solid state or other type of storage device for storing digital u-ro~ n and is connccled to data bus 18 and controlled by central proccsso~ unit 11 to store pal ~.vvol~l inforrn~tion received by speech pas~ l system 1. In addition to pas;.wold inr(,llllation, central 15 processor unit 11 is also pro~d,lulled to control voice prompt and recognition units 14 that are used to prompt users 2 and 3 in the use of speech pass~ol-l system 1 and on entering pa~wol~l inf ~rmqti-)n into the system. Typically, such prompt mPss~geS may be "Please enter your account code now.", "Please speak the passwul(l to be enrolled at this time.", and "What is your pas~wo~?" and other types of 20 mçs~o~s Referring to FIG. 2, database 10 includes a database 100 which contains ~c~.~,nce voice models ar!d related data fields, and a database 101 which contains pas;,~.ul~l files cu~ ,onding to each enrolled user. Database 100 stores a plurality of files cull~ ,onLi~g to voice entries of persons to be utilized as ler~ ce voices 25 for co. . .~ ons. In the ~ sv ,l u l ;on of database lOO, rows cc .l~;,pond to files for different l~ir~ ,nce voice contrih~tl-rs and are i-lentifiPd by a ref~-~,nce voice index (1, 2, 3...1000) co~ i"rA in the left-hand colurnn. Each person used as a reference voice speaks each of digits 0-9 and "oh", an alternative for the number zero, a ~ ~
plurality of tirnes into a ~l~cluphone such as a te~phone handset. A voice model is ~ .
30 stored for each person for each of these entries. In the illustrative elr~mrle, 1000 n,f~,lcilce voice files l~l~scn~ g different persons' voices each contain these stored ~
models. The lC.~ i.. g columns in database 100 will be described in col.junc~ion ~:
with an e~crl~n~tion of database 101.
Users seeking access through the pas~wunl system in accor~ ce with 35 the present invention, will typically be assigned an access code such as in-lir~ed in the "user" column of the table illustrative of database 101. The user is also assigned ~
21162~i~
in the illustradve tl..bodill.enl a pasOw~,ld consi~ting of five numbers defined by the "password" column. A model of each user's voice input of each assigned paO~wu~d digit is stored as a "user's word model" for each digit. As part of an initial enrollment or entry into the paO~wuld system, the user's word models of the assigned 5 pasOwc,ld are cc,lnp~;d with each of the ler.,~ ce voice models in database 100 and similarity scores are cc,l.lput~,d for digits in the stored lef~ ,nce voice model that coll~,sl.ond to digits in the pas~vvurd. The similarity scores are in-lirAAtPd by "scores"
in database 100; these scores may, for example, be on a scale from 1 tO 100 with 100 being a virtual identical match and one an e.~ cly ~ imilAr ma~ch.
As plt of the enrollment process following the r~ ;nAAtioll of these scores, each refe~nce voice file is then rank ordered by scores as in(lit AAtPd by the "rank order" column of database 100. The rank order column will consist of entries 1-1000 with 1000 coll~O~olldil~g to the rank order of the highest score (best match) and one cûll~ on~ to rank order of the lowest score (worst match).
In the illustrative ~PYAmr1e, an index is stored in each user l~aO~ rd file of database 101 to five previously i~ ;r~-ed l~;Çel~,nce voice files in database 100 selected on the basis of ranlc order. In this eYA~mrlp~ five l~r.,l~,nce voices having an a~plu~dllld~; 20 percendle rank order (200 of 1000) are selected, i.e.
indexes 1, 20, 350, 591, and 1000. These five lef~ ,nce models will be subsequently 20 utilized for paOO~. old access requests by the user as will be eYpl~i ned below.
FIG. 3 illus~rates a graph .~l~,s~,~lLing the rank ordering of the reference voice files in database 10() relative to a particular user's model of the user's assigned pasO~ .l. An il~ aspect of the present ill~,ntioll resides in the recoEnihonthat the selp~ction of comr~ri.il)n voices which are not ~.lb~l~ lu i~lly similar to the 25 voice model of the user's p~ - ~. Jld can be adv~nt~Pously utilized for validation of pdS~. old requests, especi~lly where variations in the voice rci,lJollse by a valid user or chon~inE system co~ditions can distort or change the speech cha~ ;gtic s of the user's entry of a correct pas~.ul.l. Base,d on ~p,t~ 1 results, it is believed that the s~P~ of lvf~ lce voices for co"~p~ o~c con~;de~d by rank order should be 30 selected from a group less than 80 percent (800 of 1000) in the illustrated graph and more preferably, less than 50 percent (500 of 1000) such as the illustrated range of slightly less than 20 percent (200 of 1000) to slightly greater th_n 40 percent (400 of 1000). Advantages of such selected c-""p i~;co,- voices will be filrther PYplAined below.
.
, ....
~ - 2~16~6 Referring to FIG. 4, the exemplary pasa~.c, l method begins at "START" 200 and co~ es to step 201 with the user's call being ~lsw~l~,d by password system 1. In step 203 a decision is made if the call is le4uesling enrollment, i.e. inidal entry, of a paa~wo.-l, access to pasawc,ld verification, or other 5 service not ~csoci~t~d with the pass~ l system, i.e. miccell~npous services. This d~,te. . "; n~tion can be based on the number dialed, other pdlalnel~l a controllable by F the calling party, or a DTMF reply by the calling party to a voice request generated by the system. A ~h t~ n~l;on by step 203 that neither enrollment nor access hasbeen requested, i.e. a micc~ neolls request causes the method to ~r~ in~t~ at 10 "END" 204. ~ss~lming the decision in step 203 is "enrollment", an account code or personal iflentifir~hc)n number (PIN) is assigned to the user in step 205 and it is d to the user in step 207. Next a pass~ d is assigned by the system under control of the CPU to the user in step 209. The user is l~ueat~,d to speak the ~csiened pdSa~Vv~l by a voice prompt in step 211. The user's voc~li7~tion of the15 ~csign~ pasa.~c,r~l is recorded (stored) in step 213. In step 215 a ~ e ".il.~Ron is made if the user has made an acceptable ~aponse to the request. If NO, control returns to step 211 where the user is again I~lUlll~ ,d to speak the pasa~. Jld. It will be noted that steps 211, 213, and 215 may be repeated a p~ tr ~ ---;ned number of times to store a plurality of V06 ~li7~tionc of the same paSa~ by the user in order 20 to derive a cc.,npo~ or average voc~l - on of the p8~aavvuld. A YES decision in step 215 results in a Lsconncct message being sent to the user as int~ t~d in step 217.
Continl~in~ with the paSa~.Old enrollment in FIG. 5, the CPU of ~e system 1 ~ ''e s in step 219 a co. ,~po~:le model based on the stored 2S v~li7~tion(s) of the paai~. ol'd by the user. In the illustrative eY~mrle the recorded model of the ~ . Jld numbers are then compal~;d with the voice models for cc,..~,~n~ numbers of each of the .cr~ ,nce voice files in step 221. A score is ~ccign~l to each l~f~ ,nce voice file based on similarity of the lt;r~"ence voice and user's models of col-~,s~ondillg voiced numbers irl step 223. The scores are then 30 rank ordered as ln~lic~ted in step 225. For a database cc,.~ g 1000 l~rel~,nce ~' voice files, each file would be assigned a rank between 1000 and 1 with 1000 ,~sen~ g the most similality. A pl~ le~3 number of le~,nce voices, such as five in the illustradve eY~mplç, ranks near a ~ ed target rank order, such as at a 20 percendle rank that would include ranks of 200, 201, 202, 203, and 204, in 35 step 227. The ~ank order target is preferably selected from a range of 15 - 50 perrendle and more preferably in a 20 - 40 percentile range. In step 229, the user's 211~6 . .
passwonl model and index to each selected l~;Ç~,rcllce voice are recorded in the user's file in database 101 with the assigned account code and pa~wuld. Following step 229, this method ends by exiting these steps as indif ~tPd by "END" 231.
FIG. 6 illustrates a co~ AtiQn of the flow diagram of an exemplary 5 method in accord with this invention. Upon a decision in step 2û3 of FIG.4 that "access" to the system is being ~ ei.t~,d by a user, the user is prc,.npted in step 301 (FIG. 6) for the user's account code which was assigned to the user as part of the initial pas~.ld enrollment. Step 303 rl~,~. . ,.,i,~Ps if a valid account code has been given by the user. A NO decision results in step 305 fif t ~"if.;"~ if a ~ ed 10 number of prior attempts have been eYceed~A A YES flf t~ tif~n by step 305 causes the process to be ~erfninsted as inf~ ted by "END" 306. A NO decision by step 305 which r~ se~ that the limit of tries was not f~Yf~ee~1fAfl, returns the user to step 301 for another aKempt to enter a valid account code. A YES delf ....in~lioll by step 303 col~ ,onds to the user having entered a valid account code and is followed 15 by the user being p~olll~,d to speak the user's pa~ .old at step 307. In step 309 a speech model of the pa~. old spoken by the user is saved.
In step 311 the user's file in database 101 is iA~ntifipd based upon the user's account code. The user's word rnodel stored in the id~n~ifiecl user's file is co~ d with the model of the pa~ l.l spoken by the user in le.lue~ g access, 20 and a first score is ~ n~ ,.t.,d based upon this ~ In step 313 a comr~ri~on is made of the model of the ~ - ~.Jld spoken to request access to the system with ~' each of the le~ nce voices i~l~ntifi~ by index in the user's file. A second score is t~,d based upon a c~n~ 5;~; iG of these cc", ~ ; i conS~ such as by averaging the five scores genc.~ted by these con~p~i~on~. The Lfr~ ,ncc between the first and 25 second scores is c~lrul~ted in acconldncc w~th step 315. In step 317, a ~ct .,~,;"~tion is made if the dirr~ cc d~ ~ ",;~ed in step 315 is greater than a pledct~ ",;~cdthreshold. In a typical ry~lmrle~ the first score will have a relatively high degree of similarity since the valid user's entry of the pas~wunl should relatively closely match the stored voice model of his prior entry of the same p~.Si~Wold. However, it is3û eRpect~d that the comrPri~on of the user's password with the five l~Çe~,nce voices will yield a second score indicadve of a relatively poor measure of simila~ty since each of the ref~ ,nce voices was selected to have a relative low degree of similarity, such as at the 2û percentile rank. Thus, a subst~ntial difference will typically exist between the first and second scûres~ thereby p~ lin~ a reasonable choice of 35 predele~ ed thresholds upon which to make the ~l~ t~ tion as in~iic~t~d in step 317. A threshold can be selected to provide a substantial degree of security ,~
.
;3 6 g while aecc,.. ~1~ting variadons in the voice reply of the user or of system Chi~ i.cti5C, For exarnple, a "raspy" p~ Ati~-n of the pasi,wold by a valid user such as due to a sore tl~oat may still possibly be recognized as a valid pas;,wol~l entry because it is the dirr~ ce between the first and second scores which is 5 utilized. In this eY~mple, there will be greater vaIiation and, hence a lower first score but likewise there is likely to be a greater variation with regard to the lGr~l~,nce voices and, hence a lower second score is also likely. Thus, the dirr~ntial between the filst and second scores ma-~y produce a result still high enough to meet thethreshold.
Upon a YES dete~ ;nAtion by step 317, the user is granted system access as in~lir~tç~l in step 319 and this method ~ Alf s as in~ ied by "END" 321. A NO de~ lioll by step 317 results in a ~k ~,....;I~Ation by step 323 as to whether the difference between the first and second scores ç~rcee~lfed at least a lower threshold of limit. If the lower threshold is not exceeded, i.e. a NO
15 det~ t;Ol~, the method t~rmin~tes at "END" 325. A YES ~ ....;nAti~,n by step 323 in-licpting that the lower threshold has been e-~eedeA, the method returns to step 307 and permits the user another attempt to enter the pasi,~.orll. This allows the user at least an a~ ition~l attempt to enter the pasi,wold where~at least the threshold of step 323 was met.
It is believed that the advantages of the present invention can be asc~~ from the foregoing description of an ~ f.~l of the invendon. The present invention permits a sllhst~ntiAl degree of security to be IIIAi~ ~ while also p~ g speech valialicns of the entry or l~ m;s~ n of a pas~ d to be a~cc ~ t~ d by the .~,cGglulion of the entry as a valid password. An i~ul~ol ~nt25 aspect that contribut~-s to this improved result is the sÇlpction of lG~l~,ncG voice models as the basis of a co~ .." ;con which are not subi,lA.~iAlly similar to the pdSi,~.'Of~. In the c-..ho~ of the invention, utilizing scores based upon a c~...,l-h~ ;cOi~ of the pdss..~ld entered by a user to gain access with both a stored prior entry of the passwof~l by the user and with the pl.,s~rlect~;l reference voices permits 30 a~lu~liate thresholds to be set which expand the tol ~Ice of this system as -~
CO~ A with pasi,~.ord systems in which comrArissnc with closely similar models are utilized. ' ;
Although an eA11700(1illlell~ of the present invention and a method in ~ -~
acco.d~ce ~Le.~ h have been deserihed above and illustrated in the drawings, the35 scope of the hl~,ntion is defined by the claims which follow. ~ ~
~.
Each access, line and trunk circuit 15, 16, and 17, respectively, is also coupled to switch 13 and controlled by data bus 18 eYt~n.1ing from central processor 10 unit 11. A number of voice prompt and reco~;nilion units 14 are each cc~l-n~..d with switch 13 and controlled via central proccssor unit 11 and data bus 18 to generate audio prompt me~gç~ used to instruct users 2, 3, and 5 hlt~,~u. ~nc~lrd with access, line and trunk circuits 15, 16, and 17 in the use of pas~w~,lJ i(lentific2tion system 1 and to prompt calling users 2, 3, and 5 forpa~ .Jl-l inÇo ",~1ion. Each voice prompt 15 and recognition unit 14, which is controlled by central ~OGeSSOf unit 11, llansn~
pa;,;,wc.lJ inr .""~lion received from users 2, 3, and 5 in l.,~ons~ to the gçn~r;~ted audio prompt " ~ cc ~g~ ~ over data bus 1 8 to central ~,cessol unit 11. Received pa ,;,~.o~ inf .", ~I;on is recorded under control of central processvr unit 11 in database 10. Switch 11, which is in~"lcon"~t~,d with access, line and trunk 20 circuits 15, 16, and 17, and with voice prompt and recognition units 14, is coupled to central plvces;,~l unit 11 1~ia data bus 18 and selectively connects ones of voice prompt and leco~ ion units 14 with ones of access, line and trunk circuits 15, 16, and 17, co~ e~t~A with users 2, 3, and 5.
l?asi~.Jl'l:l system 1 may co,,~l~l;ce a general purpose IBM compatible 25 C~"~ . such as a 386 or 486 or an AT&T 3B2-400 or 3B2-310 simplex or duplexco ~l.ut~, . Such Cc,~ Jut~ need not be rlescribed in detail for an ulld~ dillg of the invention and in general have a central ~ocessor unit 11 and a memory unit 12 each int~o.~l-Fct~ d by address, data and control leads to data bus 18. Data bus 18 enables central plvcessor unit 11 to co,l~""i~ir~tlo with each connected element in the 30 operationofpas~.ordsysteml. Centralpl~essolunitllisprQ~ulllllcdin ;~
acco~ ce w-th programs stored in memory 12 to enroll, recogni7e, and validate user spoken pas;,vvwJ~. ;
~I~.f~rrin~ to FIGS. 1 and 2, the illustrative speech password system 1 -records and stores in database 100 included in database 10 a plurality of speech35 models co~l~,spon(ling to strings of predefined spoken numerals entered by persons into speech pas;,.. ~ "1 system 1. Although only spoken number models are used in ..
.
2 ~ 5 6 s the illustrative embodilncnt, alph~,-". . ,- . ;r words and phrases could be used to provide a wider selection of passwords that car be assigned to users. Typically, a substûnti~l number of persons, such as 1000 persons, each record enough speech tO
allow one speech model of each of the predefined terms (numbers 0-9 and "oh") S from microphones of a trle~hc.n~ The models may be hidden markov, word models and preferably l~ c~7e.lt at least six ,t;~!lcse:~lul;on~ of each term such as based on six repetitions of each digit by each ,~Ç~.~nce voice contributor. Each model has states symbolically ~ nled as binary configurations and are preferably a comhin:-tion of multivariate Go~ n distribudons of a 24 ll;",~ n~l vector of 12 cepstral 10 coefficie~tc and 12 delta cepstral co~ffiri~nt~
Database 10 may be any one of well known disk, tape, solid state or other type of storage device for storing digital u-ro~ n and is connccled to data bus 18 and controlled by central proccsso~ unit 11 to store pal ~.vvol~l inforrn~tion received by speech pas~ l system 1. In addition to pas;.wold inr(,llllation, central 15 processor unit 11 is also pro~d,lulled to control voice prompt and recognition units 14 that are used to prompt users 2 and 3 in the use of speech pass~ol-l system 1 and on entering pa~wol~l inf ~rmqti-)n into the system. Typically, such prompt mPss~geS may be "Please enter your account code now.", "Please speak the passwul(l to be enrolled at this time.", and "What is your pas~wo~?" and other types of 20 mçs~o~s Referring to FIG. 2, database 10 includes a database 100 which contains ~c~.~,nce voice models ar!d related data fields, and a database 101 which contains pas;,~.ul~l files cu~ ,onding to each enrolled user. Database 100 stores a plurality of files cull~ ,onLi~g to voice entries of persons to be utilized as ler~ ce voices 25 for co. . .~ ons. In the ~ sv ,l u l ;on of database lOO, rows cc .l~;,pond to files for different l~ir~ ,nce voice contrih~tl-rs and are i-lentifiPd by a ref~-~,nce voice index (1, 2, 3...1000) co~ i"rA in the left-hand colurnn. Each person used as a reference voice speaks each of digits 0-9 and "oh", an alternative for the number zero, a ~ ~
plurality of tirnes into a ~l~cluphone such as a te~phone handset. A voice model is ~ .
30 stored for each person for each of these entries. In the illustrative elr~mrle, 1000 n,f~,lcilce voice files l~l~scn~ g different persons' voices each contain these stored ~
models. The lC.~ i.. g columns in database 100 will be described in col.junc~ion ~:
with an e~crl~n~tion of database 101.
Users seeking access through the pas~wunl system in accor~ ce with 35 the present invention, will typically be assigned an access code such as in-lir~ed in the "user" column of the table illustrative of database 101. The user is also assigned ~
21162~i~
in the illustradve tl..bodill.enl a pasOw~,ld consi~ting of five numbers defined by the "password" column. A model of each user's voice input of each assigned paO~wu~d digit is stored as a "user's word model" for each digit. As part of an initial enrollment or entry into the paO~wuld system, the user's word models of the assigned 5 pasOwc,ld are cc,lnp~;d with each of the ler.,~ ce voice models in database 100 and similarity scores are cc,l.lput~,d for digits in the stored lef~ ,nce voice model that coll~,sl.ond to digits in the pas~vvurd. The similarity scores are in-lirAAtPd by "scores"
in database 100; these scores may, for example, be on a scale from 1 tO 100 with 100 being a virtual identical match and one an e.~ cly ~ imilAr ma~ch.
As plt of the enrollment process following the r~ ;nAAtioll of these scores, each refe~nce voice file is then rank ordered by scores as in(lit AAtPd by the "rank order" column of database 100. The rank order column will consist of entries 1-1000 with 1000 coll~O~olldil~g to the rank order of the highest score (best match) and one cûll~ on~ to rank order of the lowest score (worst match).
In the illustrative ~PYAmr1e, an index is stored in each user l~aO~ rd file of database 101 to five previously i~ ;r~-ed l~;Çel~,nce voice files in database 100 selected on the basis of ranlc order. In this eYA~mrlp~ five l~r.,l~,nce voices having an a~plu~dllld~; 20 percendle rank order (200 of 1000) are selected, i.e.
indexes 1, 20, 350, 591, and 1000. These five lef~ ,nce models will be subsequently 20 utilized for paOO~. old access requests by the user as will be eYpl~i ned below.
FIG. 3 illus~rates a graph .~l~,s~,~lLing the rank ordering of the reference voice files in database 10() relative to a particular user's model of the user's assigned pasO~ .l. An il~ aspect of the present ill~,ntioll resides in the recoEnihonthat the selp~ction of comr~ri.il)n voices which are not ~.lb~l~ lu i~lly similar to the 25 voice model of the user's p~ - ~. Jld can be adv~nt~Pously utilized for validation of pdS~. old requests, especi~lly where variations in the voice rci,lJollse by a valid user or chon~inE system co~ditions can distort or change the speech cha~ ;gtic s of the user's entry of a correct pas~.ul.l. Base,d on ~p,t~ 1 results, it is believed that the s~P~ of lvf~ lce voices for co"~p~ o~c con~;de~d by rank order should be 30 selected from a group less than 80 percent (800 of 1000) in the illustrated graph and more preferably, less than 50 percent (500 of 1000) such as the illustrated range of slightly less than 20 percent (200 of 1000) to slightly greater th_n 40 percent (400 of 1000). Advantages of such selected c-""p i~;co,- voices will be filrther PYplAined below.
.
, ....
~ - 2~16~6 Referring to FIG. 4, the exemplary pasa~.c, l method begins at "START" 200 and co~ es to step 201 with the user's call being ~lsw~l~,d by password system 1. In step 203 a decision is made if the call is le4uesling enrollment, i.e. inidal entry, of a paa~wo.-l, access to pasawc,ld verification, or other 5 service not ~csoci~t~d with the pass~ l system, i.e. miccell~npous services. This d~,te. . "; n~tion can be based on the number dialed, other pdlalnel~l a controllable by F the calling party, or a DTMF reply by the calling party to a voice request generated by the system. A ~h t~ n~l;on by step 203 that neither enrollment nor access hasbeen requested, i.e. a micc~ neolls request causes the method to ~r~ in~t~ at 10 "END" 204. ~ss~lming the decision in step 203 is "enrollment", an account code or personal iflentifir~hc)n number (PIN) is assigned to the user in step 205 and it is d to the user in step 207. Next a pass~ d is assigned by the system under control of the CPU to the user in step 209. The user is l~ueat~,d to speak the ~csiened pdSa~Vv~l by a voice prompt in step 211. The user's voc~li7~tion of the15 ~csign~ pasa.~c,r~l is recorded (stored) in step 213. In step 215 a ~ e ".il.~Ron is made if the user has made an acceptable ~aponse to the request. If NO, control returns to step 211 where the user is again I~lUlll~ ,d to speak the pasa~. Jld. It will be noted that steps 211, 213, and 215 may be repeated a p~ tr ~ ---;ned number of times to store a plurality of V06 ~li7~tionc of the same paSa~ by the user in order 20 to derive a cc.,npo~ or average voc~l - on of the p8~aavvuld. A YES decision in step 215 results in a Lsconncct message being sent to the user as int~ t~d in step 217.
Continl~in~ with the paSa~.Old enrollment in FIG. 5, the CPU of ~e system 1 ~ ''e s in step 219 a co. ,~po~:le model based on the stored 2S v~li7~tion(s) of the paai~. ol'd by the user. In the illustrative eY~mrle the recorded model of the ~ . Jld numbers are then compal~;d with the voice models for cc,..~,~n~ numbers of each of the .cr~ ,nce voice files in step 221. A score is ~ccign~l to each l~f~ ,nce voice file based on similarity of the lt;r~"ence voice and user's models of col-~,s~ondillg voiced numbers irl step 223. The scores are then 30 rank ordered as ln~lic~ted in step 225. For a database cc,.~ g 1000 l~rel~,nce ~' voice files, each file would be assigned a rank between 1000 and 1 with 1000 ,~sen~ g the most similality. A pl~ le~3 number of le~,nce voices, such as five in the illustradve eY~mplç, ranks near a ~ ed target rank order, such as at a 20 percendle rank that would include ranks of 200, 201, 202, 203, and 204, in 35 step 227. The ~ank order target is preferably selected from a range of 15 - 50 perrendle and more preferably in a 20 - 40 percentile range. In step 229, the user's 211~6 . .
passwonl model and index to each selected l~;Ç~,rcllce voice are recorded in the user's file in database 101 with the assigned account code and pa~wuld. Following step 229, this method ends by exiting these steps as indif ~tPd by "END" 231.
FIG. 6 illustrates a co~ AtiQn of the flow diagram of an exemplary 5 method in accord with this invention. Upon a decision in step 2û3 of FIG.4 that "access" to the system is being ~ ei.t~,d by a user, the user is prc,.npted in step 301 (FIG. 6) for the user's account code which was assigned to the user as part of the initial pas~.ld enrollment. Step 303 rl~,~. . ,.,i,~Ps if a valid account code has been given by the user. A NO decision results in step 305 fif t ~"if.;"~ if a ~ ed 10 number of prior attempts have been eYceed~A A YES flf t~ tif~n by step 305 causes the process to be ~erfninsted as inf~ ted by "END" 306. A NO decision by step 305 which r~ se~ that the limit of tries was not f~Yf~ee~1fAfl, returns the user to step 301 for another aKempt to enter a valid account code. A YES delf ....in~lioll by step 303 col~ ,onds to the user having entered a valid account code and is followed 15 by the user being p~olll~,d to speak the user's pa~ .old at step 307. In step 309 a speech model of the pa~. old spoken by the user is saved.
In step 311 the user's file in database 101 is iA~ntifipd based upon the user's account code. The user's word rnodel stored in the id~n~ifiecl user's file is co~ d with the model of the pa~ l.l spoken by the user in le.lue~ g access, 20 and a first score is ~ n~ ,.t.,d based upon this ~ In step 313 a comr~ri~on is made of the model of the ~ - ~.Jld spoken to request access to the system with ~' each of the le~ nce voices i~l~ntifi~ by index in the user's file. A second score is t~,d based upon a c~n~ 5;~; iG of these cc", ~ ; i conS~ such as by averaging the five scores genc.~ted by these con~p~i~on~. The Lfr~ ,ncc between the first and 25 second scores is c~lrul~ted in acconldncc w~th step 315. In step 317, a ~ct .,~,;"~tion is made if the dirr~ cc d~ ~ ",;~ed in step 315 is greater than a pledct~ ",;~cdthreshold. In a typical ry~lmrle~ the first score will have a relatively high degree of similarity since the valid user's entry of the pas~wunl should relatively closely match the stored voice model of his prior entry of the same p~.Si~Wold. However, it is3û eRpect~d that the comrPri~on of the user's password with the five l~Çe~,nce voices will yield a second score indicadve of a relatively poor measure of simila~ty since each of the ref~ ,nce voices was selected to have a relative low degree of similarity, such as at the 2û percentile rank. Thus, a subst~ntial difference will typically exist between the first and second scûres~ thereby p~ lin~ a reasonable choice of 35 predele~ ed thresholds upon which to make the ~l~ t~ tion as in~iic~t~d in step 317. A threshold can be selected to provide a substantial degree of security ,~
.
;3 6 g while aecc,.. ~1~ting variadons in the voice reply of the user or of system Chi~ i.cti5C, For exarnple, a "raspy" p~ Ati~-n of the pasi,wold by a valid user such as due to a sore tl~oat may still possibly be recognized as a valid pas;,wol~l entry because it is the dirr~ ce between the first and second scores which is 5 utilized. In this eY~mple, there will be greater vaIiation and, hence a lower first score but likewise there is likely to be a greater variation with regard to the lGr~l~,nce voices and, hence a lower second score is also likely. Thus, the dirr~ntial between the filst and second scores ma-~y produce a result still high enough to meet thethreshold.
Upon a YES dete~ ;nAtion by step 317, the user is granted system access as in~lir~tç~l in step 319 and this method ~ Alf s as in~ ied by "END" 321. A NO de~ lioll by step 317 results in a ~k ~,....;I~Ation by step 323 as to whether the difference between the first and second scores ç~rcee~lfed at least a lower threshold of limit. If the lower threshold is not exceeded, i.e. a NO
15 det~ t;Ol~, the method t~rmin~tes at "END" 325. A YES ~ ....;nAti~,n by step 323 in-licpting that the lower threshold has been e-~eedeA, the method returns to step 307 and permits the user another attempt to enter the pasi,~.orll. This allows the user at least an a~ ition~l attempt to enter the pasi,wold where~at least the threshold of step 323 was met.
It is believed that the advantages of the present invention can be asc~~ from the foregoing description of an ~ f.~l of the invendon. The present invention permits a sllhst~ntiAl degree of security to be IIIAi~ ~ while also p~ g speech valialicns of the entry or l~ m;s~ n of a pas~ d to be a~cc ~ t~ d by the .~,cGglulion of the entry as a valid password. An i~ul~ol ~nt25 aspect that contribut~-s to this improved result is the sÇlpction of lG~l~,ncG voice models as the basis of a co~ .." ;con which are not subi,lA.~iAlly similar to the pdSi,~.'Of~. In the c-..ho~ of the invention, utilizing scores based upon a c~...,l-h~ ;cOi~ of the pdss..~ld entered by a user to gain access with both a stored prior entry of the passwof~l by the user and with the pl.,s~rlect~;l reference voices permits 30 a~lu~liate thresholds to be set which expand the tol ~Ice of this system as -~
CO~ A with pasi,~.ord systems in which comrArissnc with closely similar models are utilized. ' ;
Although an eA11700(1illlell~ of the present invention and a method in ~ -~
acco.d~ce ~Le.~ h have been deserihed above and illustrated in the drawings, the35 scope of the hl~,ntion is defined by the claims which follow. ~ ~
~.
Claims (21)
1. A spoken password verification apparatus comprising:
means for storing a first speech model of a first entry of a spoken by a user;
means for storing a plurality of reference voice (RV) speech models based on speech inputs by persons, each RV speech model including said password; means for selecting at least one RV speech model based on a predetermined difference of similarity between said first speech model and said one RV speech model;
means for storing a second speech model of a second entry of a password spoken by a user, means for comparing said second speech model and said selected one RV model to determine a measure of similarity;
means for determining the validity of said second entry of the password based on said measure of similarity.
means for storing a first speech model of a first entry of a spoken by a user;
means for storing a plurality of reference voice (RV) speech models based on speech inputs by persons, each RV speech model including said password; means for selecting at least one RV speech model based on a predetermined difference of similarity between said first speech model and said one RV speech model;
means for storing a second speech model of a second entry of a password spoken by a user, means for comparing said second speech model and said selected one RV model to determine a measure of similarity;
means for determining the validity of said second entry of the password based on said measure of similarity.
2. The spoken password verification apparatus of claim 1 wherein said means for storing RV speech models further comprises a database having files that each store an RV speech model corresponding to a person having spoken at least said password.
3. The spoken password verification apparatus of claim 1 wherein said means for selecting said at least one RV speech model comprises means for determining a similarity score for each RV speech model by comparing said first speech model with each of said RV speech models.
4. The spoken password verification apparatus of claim 3 wherein said means for selecting said at least one RV speech model further comprises means for ranking said RV speech models based on said similarity scores, said selecting means selecting said at least one RV speech model based on a predetermined threshold ranking that corresponds to a substantial difference of similarity scores of said first speech model and said at least one RV speech model.
5. The spoken password verification apparatus of claim 3 wherein said means for selecting said at least one RV speech model further comprises means for ranking said RV speech models based on said similarity scores, said selecting means selecting said at least one RV speech model based on a predetermined threshold ranking from a range of said RV speech model rankings of 15 - 50 percentile, where a 100 percentile ranking represents an RV speech model with the most similarity to said first speech model.
6. The spoken password verification apparatus of claim 5 wherein said range of said RV speech model rankings consists of 20 - 40 percentile.
7. The spoken password verification apparatus of claim 1 wherein said determining means comprises means for generating first and second similarity scores based on a comparison of said second speech model with said first speech model and with said at least one RV speech model, respectively, means for comparing the difference between said first and second similarity scores, and said determiningmeans determining that the password corresponding to said second speech model isvalid if said difference between said first and second similarity scores exceeds a predetermined threshold.
8. A method for verifying a spoken password comprising the steps of:
storing a first speech model of a first entry of a password spoken by a user;
storing a plurality of reference voice (RV) speech models based on speech inputs by persons, each RV speech model including said password;
selecting at least one RV speech model based on a predetermined difference of similarity between said first speech model and said one RV speech model;
storing a second speech model of a second entry of a password spoken by a user;
comparing said second speech model and said selected one RV model to determine a measure of similarity;
determining the validity of said second entry of the password based on said measure of similarity.
storing a first speech model of a first entry of a password spoken by a user;
storing a plurality of reference voice (RV) speech models based on speech inputs by persons, each RV speech model including said password;
selecting at least one RV speech model based on a predetermined difference of similarity between said first speech model and said one RV speech model;
storing a second speech model of a second entry of a password spoken by a user;
comparing said second speech model and said selected one RV model to determine a measure of similarity;
determining the validity of said second entry of the password based on said measure of similarity.
9. The spoken password verification method of claim 8 wherein said step of storing RV speech models further comprises storing said RV speech models in adatabase having files that each store an RV speech model corresponding to a person having spoken at least said password.
10. The spoken password verification method of claim 8 wherein said step of selecting said at least one RV speech model comprises the step of determining a similarity score for each RV speech model by comparing said first speech model with each of said RV speech models.
11. The spoken password verification method of claim 10 wherein said step of selecting said at least one RV speech model further comprises the step of ranking said RV speech models based on said similarity scores, said selecting step selecting said at least one RV speech model based on a predetermined threshold ranking that corresponds to a substantial difference of similarity scores of said first speech model and said at least one RV speech model.
12. The spoken password verification method of claim 10 wherein said step of selecting said at least one RV speech model further comprises the step of ranking said RV speech models based on said similarity scores, said selecting step selecting said at least one RV speech model based on a predetermined threshold ranking from a range of said RV speech model rankings of 15 - 50 percentile, where a 100 percentile ranking represents an RV speech model with the most similarity to said first speech model.
13. The spoken password verification method of claim 12 wherein said range of said RV speech model rankings consists of 20 - 40 percentile.
14. The spoken password verification method of claim 8 wherein said determining step comprises the steps of generating first and second similarity scores based on a comparison of said second speech model with said first speech model and with said at least one RV speech model, respectively, comparing the difference between said first and second similarity scores, and determining that the password corresponding to said second speech model is valid if said difference between said first and second similarity scores exceeds a predetermined threshold.
15. A password verification system comprising:
database:
first speech model of a first entry of a password spoken by a stored in said database;
a plurality of reference voice (RV) speech models based on speech inputs by persons stored in said database, each RV speech model including said password;
means operating under the control of a computer program for selecting at least one RV speech model based on a predetermined difference of similarity between said first speech model and said one RV speech model;
a second speech model of a second entry of a password spoken by a user stored in said database;
means operating under the control of a computer program for comparing said second speech model and said selected one RV model to determine a measure of similarity;
means operating under the control of a computer program for determining the validity of said second entry of the password based on said measure of similarity.
database:
first speech model of a first entry of a password spoken by a stored in said database;
a plurality of reference voice (RV) speech models based on speech inputs by persons stored in said database, each RV speech model including said password;
means operating under the control of a computer program for selecting at least one RV speech model based on a predetermined difference of similarity between said first speech model and said one RV speech model;
a second speech model of a second entry of a password spoken by a user stored in said database;
means operating under the control of a computer program for comparing said second speech model and said selected one RV model to determine a measure of similarity;
means operating under the control of a computer program for determining the validity of said second entry of the password based on said measure of similarity.
16. The password system of claim 15 wherein said database includes files that each store an RV speech model corresponding to a person having spoken at least said password.
17. The password system of claim 15 wherein said means for selecting said at least one RV speech model comprises means operating under the control of a computer program for determining a similarity score for each RV speech model by comparing said first speech model with each of said RV speech models.
18. The password system of claim 17 wherein said means for selecting said at least one RV speech model further comprises means operating under the control of a computer program for ranking said RV speech models based on said similarity scores, said selecting means selecting said at least one RV speech model based on a predetermined threshold ranking that corresponds to a substantial difference of similarity scores of said first speech model and said at least one RV
speech model.
speech model.
19. The password system of claim 17 wherein said means for selecting said at least one RV speech model further comprises means operating under the control of a computer program for ranking said RV speech models based on said similarity scores, said selecting means selecting said at least one RV speech model based on a predetermined threshold ranking from a range of said RV speech model rankings of 15 - 50 percentile, where a 100 percentile ranking represents an RV
speech model with the most similarity to said first speech model.
speech model with the most similarity to said first speech model.
20. The password system of claim 19 wherein said range of said RV
speech model rankings consists of 20 - 40 percentile.
speech model rankings consists of 20 - 40 percentile.
21. The password system of claim 15 wherein said determining means comprises means operating under the control of a computer program for generatingfirst and second similarity scores based on a comparison of said second speech model with said first speech model and with said at least one RV speech model, respectively, means operating under the control of a computer program for comparing the difference between said first and second similarity scores, and said determining means determining that the password corresponding to said second speech model is valid if said difference between said first and second similarity scores exceeds a predetermined threshold.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US052,310 | 1993-04-23 | ||
| US08/052,310 US5430827A (en) | 1993-04-23 | 1993-04-23 | Password verification system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2116256A1 CA2116256A1 (en) | 1994-10-24 |
| CA2116256C true CA2116256C (en) | 1997-09-30 |
Family
ID=21976758
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002116256A Expired - Fee Related CA2116256C (en) | 1993-04-23 | 1994-02-23 | Password verification system |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US5430827A (en) |
| EP (1) | EP0621532B1 (en) |
| CA (1) | CA2116256C (en) |
| DE (1) | DE69421310T2 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107193391A (en) * | 2017-04-25 | 2017-09-22 | 北京百度网讯科技有限公司 | The method and apparatus that a kind of upper screen shows text message |
Families Citing this family (62)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5758317A (en) * | 1993-10-04 | 1998-05-26 | Motorola, Inc. | Method for voice-based affiliation of an operator identification code to a communication unit |
| US5583933A (en) | 1994-08-05 | 1996-12-10 | Mark; Andrew R. | Method and apparatus for the secure communication of data |
| US5907597A (en) | 1994-08-05 | 1999-05-25 | Smart Tone Authentication, Inc. | Method and system for the secure communication of data |
| US5671271A (en) | 1995-05-25 | 1997-09-23 | Henderson; Daniel A. | Dialer programming system and device with integrated printing process |
| US5839103A (en) * | 1995-06-07 | 1998-11-17 | Rutgers, The State University Of New Jersey | Speaker verification system using decision fusion logic |
| US5774858A (en) * | 1995-10-23 | 1998-06-30 | Taubkin; Vladimir L. | Speech analysis method of protecting a vehicle from unauthorized accessing and controlling |
| FR2743166B1 (en) * | 1995-12-28 | 1998-01-30 | Kaspar Emmanuel Charles Albert | DATA PROCESSING DEVICE AND SYSTEM INCORPORATING A VOICE CONTROL LOOP |
| US5956491A (en) | 1996-04-01 | 1999-09-21 | Marks; Daniel L. | Group communications multiplexing system |
| US5937381A (en) * | 1996-04-10 | 1999-08-10 | Itt Defense, Inc. | System for voice verification of telephone transactions |
| SE505522C2 (en) * | 1996-07-01 | 1997-09-08 | Telia Ab | Procedure and arrangement for adapting models to, for example, speaker verification systems |
| DE19630109A1 (en) * | 1996-07-25 | 1998-01-29 | Siemens Ag | Method for speaker verification using at least one speech signal spoken by a speaker, by a computer |
| US6453281B1 (en) | 1996-07-30 | 2002-09-17 | Vxi Corporation | Portable audio database device with icon-based graphical user-interface |
| ATE227868T1 (en) * | 1996-10-15 | 2002-11-15 | Swisscom Ag | METHOD FOR VERIFYING A SPEAKER |
| US5915001A (en) * | 1996-11-14 | 1999-06-22 | Vois Corporation | System and method for providing and using universally accessible voice and speech data files |
| US5963908A (en) * | 1996-12-23 | 1999-10-05 | Intel Corporation | Secure logon to notebook or desktop computers |
| US6012027A (en) * | 1997-05-27 | 2000-01-04 | Ameritech Corporation | Criteria for usable repetitions of an utterance during speech reference enrollment |
| US7630895B2 (en) * | 2000-01-21 | 2009-12-08 | At&T Intellectual Property I, L.P. | Speaker verification method |
| AU8125198A (en) * | 1997-06-27 | 1999-01-19 | Lernout & Hauspie Speech Products N.V. | Access-controlled computer system with automatic speech recognition |
| US6014650A (en) * | 1997-08-19 | 2000-01-11 | Zampese; David | Purchase management system and method |
| US6026491A (en) * | 1997-09-30 | 2000-02-15 | Compaq Computer Corporation | Challenge/response security architecture with fuzzy recognition of long passwords |
| US6246987B1 (en) * | 1998-02-04 | 2001-06-12 | Alcatel Usa Sourcing, L.P. | System for permitting access to a common resource in response to speaker identification and verification |
| WO1999042992A1 (en) * | 1998-02-24 | 1999-08-26 | Holoubek Michael J | Randomly generated voice recognition method and apparatus |
| US6185536B1 (en) * | 1998-03-04 | 2001-02-06 | Motorola, Inc. | System and method for establishing a communication link using user-specific voice data parameters as a user discriminator |
| US6618806B1 (en) | 1998-04-01 | 2003-09-09 | Saflink Corporation | System and method for authenticating users in a computer network |
| US6928547B2 (en) * | 1998-07-06 | 2005-08-09 | Saflink Corporation | System and method for authenticating users in a computer network |
| US8212893B2 (en) | 1999-06-08 | 2012-07-03 | Verisign, Inc. | Digital camera device and methodology for distributed processing and wireless transmission of digital images |
| US6473735B1 (en) * | 1999-10-21 | 2002-10-29 | Sony Corporation | System and method for speech verification using a confidence measure |
| WO2001069543A1 (en) * | 2000-03-14 | 2001-09-20 | British Telecommunications Public Limited Company | Secure services |
| EP1134703A1 (en) * | 2000-03-14 | 2001-09-19 | BRITISH TELECOMMUNICATIONS public limited company | Secure services |
| EP2278538A1 (en) * | 2000-04-24 | 2011-01-26 | Visa International Service Association | Online payer authentication service |
| US20040215788A1 (en) * | 2000-07-26 | 2004-10-28 | Morris Robert P. | System and method for associating a device with an existing service account |
| US8224776B1 (en) | 2000-07-26 | 2012-07-17 | Kdl Scan Designs Llc | Method and system for hosting entity-specific photo-sharing websites for entity-specific digital cameras |
| US6636259B1 (en) * | 2000-07-26 | 2003-10-21 | Ipac Acquisition Subsidiary I, Llc | Automatically configuring a web-enabled digital camera to access the internet |
| US7287088B1 (en) * | 2000-10-06 | 2007-10-23 | Fotomedia Technologies, Llc | Transmission bandwidth and memory requirements reduction in a portable image capture device by eliminating duplicate image transmissions |
| US6934852B2 (en) * | 2000-12-11 | 2005-08-23 | International Business Machines Corporation | Security keys for enhanced downstream access security for electronic file systems and drives |
| US7197531B2 (en) * | 2000-12-29 | 2007-03-27 | Fotomedia Technologies, Llc | Meta-application architecture for integrating photo-service websites for browser-enabled devices |
| US7272788B2 (en) * | 2000-12-29 | 2007-09-18 | Fotomedia Technologies, Llc | Client-server system for merging of metadata with images |
| US7433710B2 (en) * | 2001-04-20 | 2008-10-07 | Lightsurf Technologies, Inc. | System and methodology for automated provisioning of new user accounts |
| US20020010715A1 (en) * | 2001-07-26 | 2002-01-24 | Garry Chinn | System and method for browsing using a limited display device |
| JP2004537116A (en) * | 2001-07-27 | 2004-12-09 | マルティン セルゲーヴィッチ ヌヌパロフ | Method and apparatus for inputting password for accessing computer database |
| US20040015243A1 (en) * | 2001-09-28 | 2004-01-22 | Dwyane Mercredi | Biometric authentication |
| US7724281B2 (en) * | 2002-02-04 | 2010-05-25 | Syniverse Icx Corporation | Device facilitating efficient transfer of digital content from media capture device |
| US20040015702A1 (en) * | 2002-03-01 | 2004-01-22 | Dwayne Mercredi | User login delegation |
| US20030171930A1 (en) * | 2002-03-07 | 2003-09-11 | Junqua Jean-Claude | Computer telephony system to access secure resources |
| US7064652B2 (en) * | 2002-09-09 | 2006-06-20 | Matsushita Electric Industrial Co., Ltd. | Multimodal concierge for secure and convenient access to a home or building |
| US20040059590A1 (en) * | 2002-09-13 | 2004-03-25 | Dwayne Mercredi | Credential promotion |
| US7360694B2 (en) * | 2003-01-23 | 2008-04-22 | Mastercard International Incorporated | System and method for secure telephone and computer transactions using voice authentication |
| GB0304472D0 (en) * | 2003-02-27 | 2003-04-02 | Impressive Ideas Ltd | Remote access system for digital data storage media |
| US7467403B2 (en) * | 2004-01-09 | 2008-12-16 | Harris Scott C | Techniques for entry of less-than-perfect-passwords |
| US7373516B2 (en) | 2004-08-19 | 2008-05-13 | International Business Machines Corporation | Systems and methods of securing resources through passwords |
| US20060136741A1 (en) * | 2004-12-16 | 2006-06-22 | Saflink Corporation | Two factor token identification |
| US20060143607A1 (en) * | 2004-12-29 | 2006-06-29 | Morris Robert P | Method and system for allowing a user to create actions to be taken by a server |
| US20060143684A1 (en) * | 2004-12-29 | 2006-06-29 | Morris Robert P | Method and system for allowing a user to specify actions that are to be automatically performed on data objects uploaded to a server |
| US20060293898A1 (en) * | 2005-06-22 | 2006-12-28 | Microsoft Corporation | Speech recognition system for secure information |
| US20070219792A1 (en) * | 2006-03-20 | 2007-09-20 | Nu Echo Inc. | Method and system for user authentication based on speech recognition and knowledge questions |
| US8108932B2 (en) * | 2008-06-12 | 2012-01-31 | International Business Machines Corporation | Calculating a password strength score based upon character proximity and relative position upon an input device |
| CN101997849A (en) * | 2009-08-18 | 2011-03-30 | 阿里巴巴集团控股有限公司 | Method, device and system for authentication of internet user |
| WO2012017384A1 (en) | 2010-08-02 | 2012-02-09 | 3Fish Limited | Identity assessment method and system |
| US20120155663A1 (en) * | 2010-12-16 | 2012-06-21 | Nice Systems Ltd. | Fast speaker hunting in lawful interception systems |
| US9679496B2 (en) | 2011-12-01 | 2017-06-13 | Arkady Zilberman | Reverse language resonance systems and methods for foreign language acquisition |
| US9864576B1 (en) * | 2013-09-09 | 2018-01-09 | Amazon Technologies, Inc. | Voice controlled assistant with non-verbal user input |
| US10909240B2 (en) * | 2017-12-18 | 2021-02-02 | Johnson Controls Technology Company | Building management system with malicious user detection and prevention |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4053710A (en) * | 1976-03-01 | 1977-10-11 | Ncr Corporation | Automatic speaker verification systems employing moment invariants |
| US4363102A (en) * | 1981-03-27 | 1982-12-07 | Bell Telephone Laboratories, Incorporated | Speaker identification system using word recognition templates |
| JPS5876893A (en) * | 1981-10-30 | 1983-05-10 | 日本電気株式会社 | Voice recognition equipment |
| JPS59178587A (en) * | 1983-03-30 | 1984-10-09 | Nec Corp | Speaker confirming system |
| US4910782A (en) * | 1986-05-23 | 1990-03-20 | Nec Corporation | Speaker verification system |
| EP0444351A3 (en) * | 1990-02-28 | 1992-05-27 | American Telephone And Telegraph Company | Voice password-controlled computer security system |
| US5127043A (en) * | 1990-05-15 | 1992-06-30 | Vcs Industries, Inc. | Simultaneous speaker-independent voice recognition and verification over a telephone network |
-
1993
- 1993-04-23 US US08/052,310 patent/US5430827A/en not_active Expired - Lifetime
-
1994
- 1994-02-23 CA CA002116256A patent/CA2116256C/en not_active Expired - Fee Related
- 1994-04-13 DE DE69421310T patent/DE69421310T2/en not_active Expired - Lifetime
- 1994-04-13 EP EP94302593A patent/EP0621532B1/en not_active Expired - Lifetime
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107193391A (en) * | 2017-04-25 | 2017-09-22 | 北京百度网讯科技有限公司 | The method and apparatus that a kind of upper screen shows text message |
Also Published As
| Publication number | Publication date |
|---|---|
| EP0621532B1 (en) | 1999-10-27 |
| DE69421310T2 (en) | 2000-07-13 |
| CA2116256A1 (en) | 1994-10-24 |
| EP0621532A1 (en) | 1994-10-26 |
| DE69421310D1 (en) | 1999-12-02 |
| US5430827A (en) | 1995-07-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2116256C (en) | Password verification system | |
| US5806040A (en) | Speed controlled telephone credit card verification system | |
| EP0686297B1 (en) | Voice command control and verification system and method | |
| US6119084A (en) | Adaptive speaker verification apparatus and method including alternative access control | |
| US5216720A (en) | Voice verification circuit for validating the identity of telephone calling card customers | |
| US6556127B1 (en) | Speaker verification method | |
| US6076055A (en) | Speaker verification method | |
| US6012027A (en) | Criteria for usable repetitions of an utterance during speech reference enrollment | |
| US5054083A (en) | Voice verification circuit for validating the identity of an unknown person | |
| US20030074201A1 (en) | Continuous authentication of the identity of a speaker | |
| CA2104850C (en) | Speech password system | |
| US9373325B2 (en) | Method of accessing a dial-up service | |
| US6246987B1 (en) | System for permitting access to a common resource in response to speaker identification and verification | |
| CA2040724A1 (en) | Voice password controlled computer security system | |
| US20080071538A1 (en) | Speaker verification method | |
| AU2385700A (en) | Security and user convenience through voice commands | |
| US9978373B2 (en) | Method of accessing a dial-up service | |
| Fogel | A Commercial Implementation of a Free-Speech Speaker Verification System in a Call Center |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKLA | Lapsed |