AU2012205747B9 - Natural enrolment process for speaker recognition - Google Patents

Natural enrolment process for speaker recognition Download PDF

Info

Publication number
AU2012205747B9
AU2012205747B9 AU2012205747A AU2012205747A AU2012205747B9 AU 2012205747 B9 AU2012205747 B9 AU 2012205747B9 AU 2012205747 A AU2012205747 A AU 2012205747A AU 2012205747 A AU2012205747 A AU 2012205747A AU 2012205747 B9 AU2012205747 B9 AU 2012205747B9
Authority
AU
Australia
Prior art keywords
biometric
agent
user
verification
software
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.)
Active
Application number
AU2012205747A
Other versions
AU2012205747B2 (en
AU2012205747A1 (en
Inventor
Wilson A. Garcia
Garfield SAMUEL
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Validvoice LLC
Original Assignee
Validvoice LLC
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Validvoice LLC filed Critical Validvoice LLC
Publication of AU2012205747A1 publication Critical patent/AU2012205747A1/en
Publication of AU2012205747B2 publication Critical patent/AU2012205747B2/en
Application granted granted Critical
Publication of AU2012205747B9 publication Critical patent/AU2012205747B9/en
Assigned to VALIDVOICE, LLC reassignment VALIDVOICE, LLC Request for Assignment Assignors: GARCIA, WILSON, SAMUEL, Garfield
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L9/00Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
    • H04L9/32Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols including means for verifying the identity or authority of a user of the system or for message authentication, e.g. authorization, entity authentication, data integrity or data verification, non-repudiation, key authentication or verification of credentials
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L9/00Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
    • H04L9/32Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols including means for verifying the identity or authority of a user of the system or for message authentication, e.g. authorization, entity authentication, data integrity or data verification, non-repudiation, key authentication or verification of credentials
    • H04L9/3226Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols including means for verifying the identity or authority of a user of the system or for message authentication, e.g. authorization, entity authentication, data integrity or data verification, non-repudiation, key authentication or verification of credentials using a predetermined code, e.g. password, passphrase or PIN
    • H04L9/3231Biological data, e.g. fingerprint, voice or retina
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q20/00Payment architectures, schemes or protocols
    • G06Q20/38Payment protocols; Details thereof
    • G06Q20/40Authorisation, e.g. identification of payer or payee, verification of customer or shop credentials; Review and approval of payers, e.g. check credit lines or negative lists
    • G06Q20/401Transaction verification
    • G06Q20/4014Identity check for transactions
    • G06Q20/40145Biometric identity checks
    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07CTIME 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
    • G07C9/00Individual registration on entry or exit
    • G07C9/30Individual registration on entry or exit not involving the use of a pass
    • G07C9/32Individual registration on entry or exit not involving the use of a pass in combination with an identity check
    • G07C9/37Individual 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
    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07CTIME 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
    • G07C9/00Individual registration on entry or exit
    • G07C9/30Individual registration on entry or exit not involving the use of a pass
    • G07C9/38Individual registration on entry or exit not involving the use of a pass with central registration
    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L17/00Speaker identification or verification techniques
    • G10L17/04Training, enrolment or model building
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M3/00Automatic or semi-automatic exchanges
    • H04M3/42Systems providing special services or facilities to subscribers
    • H04M3/487Arrangements for providing information services, e.g. recorded voice services or time announcements
    • H04M3/493Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals
    • H04M3/4936Speech interaction details
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M3/00Automatic or semi-automatic exchanges
    • H04M3/42Systems providing special services or facilities to subscribers
    • H04M3/50Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
    • H04M3/51Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
    • H04M3/5183Call or contact centers with computer-telephony arrangements
    • GPHYSICS
    • G07CHECKING-DEVICES
    • G07CTIME 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
    • G07C2209/00Indexing scheme relating to groups G07C9/00 - G07C9/38
    • G07C2209/02Access control comprising means for the enrolment of users
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L2463/00Additional details relating to network architectures or network communication protocols for network security covered by H04L63/00
    • H04L2463/102Additional details relating to network architectures or network communication protocols for network security covered by H04L63/00 applying security measure for e-commerce
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M2201/00Electronic components, circuits, software, systems or apparatus used in telephone systems
    • H04M2201/41Electronic components, circuits, software, systems or apparatus used in telephone systems using speaker recognition
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M2203/00Aspects of automatic or semi-automatic exchanges
    • H04M2203/60Aspects of automatic or semi-automatic exchanges related to security aspects in telephonic communication systems
    • H04M2203/6054Biometric subscriber identification

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Human Computer Interaction (AREA)
  • Signal Processing (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Accounting & Taxation (AREA)
  • Multimedia (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Acoustics & Sound (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Marketing (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Finance (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Telephonic Communication Services (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Description

WO 2012/096901 PCT/US2012/020679 NATURAL ENROLMENT PROCESS FOR SPEAKER RECOGNITION U.S. Government Support [0001] N/A Description of Area and Background of the Invention [0002] 1). Technical Field [00031 The present invention relates generally to a method for biometric authentication, and more particularly, to automating and providing continuous application of algorithms during an interaction, with the use of uncertainty modeling, to minimize the inconvenience of enrolment of users and maximize security of access by securing the entire interaction. [0004] 2). Background [0005] Organizations with large constituencies dealing with private or financial information or transactions often require that the identity of the parties to a transaction be verified to ensure that they have the proper authority to receive sensitive information or complete a particular transaction. These parties may be the organization's customers or the organization's employees or agents. In the case of the organization's representatives validation of qualifications or authorization may be necessary. [0006] A range of techniques is used involving one or more criteria to deliver a degree of certainty that an i ndividual is who he c claims t o be. The criteria m ay be based on something the individual knows (birth date, password), something the individual has (a token of some kind) or a unique physical attribute (biometric data-i.e., voiceprint, iris scan, retina scan, fingerprint, etc.). [0007] Biometric verification of identity requires that first the individual concerned be enrolled so that upon verification one knows the physical attribute that one is attempting to match. Verification typically only occurs at the beginning of an interaction and once achieved is considered as valid for the whole interaction. This does not allow for instances where the validation of the individuals identity can be completed correctly, but the interaction is then taken up by another party. This is a serious potential flaw for those interactions requiring a very high standard of identity verification. [0008] Using the voice as a biometric attribute has the advantage that the enrolment and verification can be completed over common telephony systems. Also unlike passwords and tokens voice attributes cannot be forgotten or easily stolen. Furthermore, verification WO 2012/096901 PCT/US2012/020679 -2 can be repeated throughout the interaction preventing an imposter from taking over a verified interaction. However, enrolment of users is viewed as being a cumbersome and time consuming process because to achieve a reliable enrolment requires a substantial amount of speech to be captured. To date this has limited the adoption of voice biometrics as a means of telephone-based authentication. [0009] 3). Related Art [0010] U.S. Patent No. 6,078,807 to Dunn et al. discloses a system/method for screening telephone system callers, using speaker-dependent text-independent voice recognition, to either prevent excessive fraudulent use of certain telephone system lines (or radio channels) or to selectively restrict classes of service allowed for different users of a common line. The voice recognition process is configured to provide an extent of mismatch indication, between a voice sample taken from a caller currently using the protected line (or channel) and a recorded reference voice sample, which indication corresponds to the probability that the instant caller is not the same as the person who provided the reference sample. [0011] U.S. Patent No 7,212,613 to Kim et al. discloses a voice biometric authentication system and method wherein an authentication system is provided for authenticating a user of a telephonic device. The system includes a setup system for capturing and storing an authentic user voice sample, a comparison system for comparing the authentic user voice sample with an inputted voice sample and generating a comparison result, and a control system for controlling access to the telephonic device. The control system analyzes the comparison result for an initial inputted voice sample received when a telephone call is initiated, and periodically reanalyzes comparison results for ongoing inputted voice samples received during the telephone call. [0012] U.S. Patent No. 7,340,042 to Cluff et al. discloses a system and method of providing voice authentication to institutional customers having consumers or users. The system includes an enrollment module, a verification module, and a database. Each institutional customer has an authentication profile on the system that includes both instructions indicating which one or more of the voice biometric engines are to be used for a given consumer's enrollment and/or authentication, and also variables for implementing those instructions. A given institutional customer may have more than one set of instructions or variables for different classes of users. Vocal information is received from a user, and a baseline voiceprint is constructed for enrollment. Subsequently, more vocal information is received from the user to construct a temporary voiceprint. The temporary WO 2012/096901 PCT/US2012/020679 -3 voiceprint is compared to a corresponding baseline voiceprint in accordance with the institutional customer's authentication profile. [0013] U.S. Patent No. 7,367,049 to Robinson et al. discloses a system and method of enrolling potential system users for a biometric system for identity verification. The potential system user information is entered into the system, either by the user or a system operator, and is stored as a partially-enabled user record. The user of a partially enabled user record fully enables the record by presenting information previously stored in the user record and presenting the remainder of user information necessary to complete record activation. Enrollment data is used to authenticate the system user's identity and authorize related transaction accesses in a biometric system for identity verification. Summary of the Invention [0014] In accordance with various embodiments of the present invention, the described system provides the capability to continually monitor a voice interaction for the purposes of voice biometric enrolment and verification whilst providing feedback to the organizations agent of the current status. [0015] From the perspective of enrolment this allows the live agent to guide the enrollee through the process. The design ensures that the enrollee's voice only is processed for the purpose of enrolment. [0016] It is expected that an interaction guided by a live agent will have a greater chance of a successful outcome than an automated interaction. The agent will be provided with feedback to either a visual display or via a "whisper" (inaudible to the enrollee) to the agent that advises whether sufficient quantity and quality of voice data has been captured to complete enrolment. [0017] The design of the interaction dialogue and the enrollee's awareness of what is being undertaken is a matter of the organization's policies and any requirement of the local jurisdiction. [0018] From the verification perspective the design provides for continuous verification of all parties identities throughout the transaction. This will detect if during the transaction any party changes or any unverifiable party should make any utterances. From the perspective of the customer this provides enhanced security. For the organization it also allows verification that an authorized agent is undertaking the interaction. This is of WO 2012/096901 PCT/US2012/020679 -4 particular value for remotely based agents where the organization may have limited physical control. [0019] Notification of changes to the verification status would, as appropriate, be made known to the agent or supervisor or both. This could be via visual alarm or whisper or both. As an option the interaction could be terminated. [0020] Because the agent and customer voice streams are carefully separated, it is possible to undertake continuous verification of the agent identity while enrolling the customer using the natural Enrolment Process, [0021] Verification scores will be presented to the agent or supervisor or any other party either directly or further processed in accordance with the organizations policies. [0022] The preferred embodiment uses verification scores that are calculated using the techniques as outlined in copending U.S. Patent application No. 12/663,163 Al filed Dec 19, 2007 and entitled Confidence levels for Speaker Recognition but may use any other methodology. Description of the Fiqures [0023] FIGURE 1 is a block diagram illustrating one embodiment of a Natural Enrolment Process, deployed in a variety of environments. [0024] FIGURE 2 is a flow diagram of an embodiment of the inventive enrolment process wherein no agent intervention is required in the enrolment process. [0025] FIGURE 3 is a flow diagram of an embodiment of the inventive enrolment process wherein a live agent participates in the verification of a previously enrolled user aided by the enrolment widget. [0026] FIGURE 4 is a flow diagram of an embodiment of the inventive enrolment process used for verification purposes wherein the system handles the verification of a previously enrolled user (although a live agent may participate in the judgment of a questionable case). [0027] FIGURE 5 is a flow diagram of an embodiment of the inventive enrolment process used for verification purposes wherein a live agent participates in the verification of a previously enrolled user aided by the enrolment widget.
WO 2012/096901 PCT/US2012/020679 -5 Detailed Description of the Invention [0028] The detailed description set forth below in connection with the appended drawings is intended as a description of certain embodiments of the present disclosure, and is not intended to represent the only forms that may be developed or utilized. The description sets forth the various functions in connection with the illustrated embodiments, but it is to be understood, however, that the same or equivalent functions may be accomplished by different embodiments that are also intended to be encompassed within the scope of the present disclosure. It is further understood that the use of relational terms such as top and bottom, first and second, and the like are used solely to distinguish one entity from another without necessarily requiring or implying any actual such relationship or order between such entities. [0029] With reference to the block diagram of FIG. 1, one embodiment of the Natural Enrolment Process interfaces between multiple telecommunications and computing networks including, but not limited to, the Public Switched Telephone Network (PSTN), IP-based networks such as the Internet, mobile cellular networks and multimedia-oriented high speed networks including, but not limited to, cable TV, IP TV, and satellite networks. [0030] Enrollee Aspect [0031] At 100 the Enrollee Aspect is described. The Enrollee Aspect is comprised of a User 105 Terminal Device 110 such as a PSTN 115 and/or alternately a SIPNolP (voice over IP) network 120. The Enrollee Caller 105 initiates either an enrollment transaction or a verification transaction by making a phone call to a system which employs a Biometric Complex 600, 700, 800. [0032] In a preferred embodiment of the invention the User 105 initiates an enrolment or verification transaction by using a telephone instrument or Terminal Device 110; however, this device is not be limited to a telephonic device but may also include a microphone in combination with another terminal device that would forward recorded voice utterances over an IP or telephone network to the Biometric Engine 600. [0033] At 115 the user 105 makes a phone call using the Terminal Device 110 over the PSTN. Alternately the user 105 may use a VolP enabled device 110 to initiate the transaction over an IP based network. [0034] Overlay complex. [0035] The overlay complex is comprised of a VolP Gateway 125 and/or Session Border Controller 130 and an Optional Preamble Engine 200. At 125 the phone call from the WO 2012/096901 PCT/US2012/020679 -6 User 105 terminates on a common VolP gateway. Such gateways are available from vendors such as CISCO, Sangoma and Audiocodes. One of ordinary skill in the art of telephony will understand that the Gateway 125 can convert PSTN signals into VoIP signals for further transmission to an IP based telephony engine. [0036] In A preferred embodiment of the invention an IP signal would be processed by the Optional Preamble Engine 200. At 130 the IP based telephone call from the User 105 will terminate on a Session Border Controller which is a commonly available device. Such devices are available from vendors such as CISCO and ACME Packets. In the context of this invention the purpose of the Session Border Controller is to provide security and firewall functions for incoming and outgoing IP based telephone calls. [0037] At 200 any phone call from the User 105 regardless as to its conveyance from a Gateway 125 or a Session Border Controller 130 may be further processed in a preferred embodiment of the invention a simple "drop and insert" function can be achieved with the Preamble Engine. To one of ordinary skill in the art such a "drop and insert' function can be built using the apparatus described in US Patent No. 5,046,088 entitled "Converter for in-band routing and/or origination information". [0038] Here at 200 the IP based voice communication coming from the User 105 can be discretely recorded and analyzed in real time by the Biometric Complex 600, 700, 800. This establishes a method for analyzing individual voice print data whilst allowing intervention of a live agent 505 at the same time. [0039] Alternately, an Optional Preamble Engine 200 can be used as a switched resource behind a switching point 310 or 320 as pictured at 310 and 320. Here signaling schemes such as Verizon Businesses ECR or CISCO's ICM can be used to share Preamble Engine resources only during those portions of the call where the Biometric Complex needs to be evoked. [0040] Native Complex. [0041] At 210 the original phone call propagates from the Optional Preamble Engine 200 and continues to the Primary Preamble System 300. This Primary Preamble System may be an existing i nteractive voice response (IVR) system either collocated or network based. In a preferred embodiment of the invention the Optional Preamble Engine 200 along with its ability to record and analyze the User 105 voice in real time would be combined with the Primary PS 300. However it may be advantageous for the preamble engine 200 to stand alone if an existing IVR system cannot be changed.
WO 2012/096901 PCT/US2012/020679 -7 [0042] Likewise at 220 the original IP based call continues to propagate into the Primary Preamble System 300 in the same fashion as the PSTN based call. [0043] At 310 the original PSTN phone call exits the Primary Preamble System 300 and terminates on auto call distributor/private branch exchange (ACD/PABX) 400. Such a transfer between Primary Preamble System 300 and ACD/PABX 400 is well known by one of ordinary skill in the art of contact center systems. For example, this transfer may happen using regular telephony signaling using a direct connection between the Primary Preamble System 300 and the ACD/PABX 400. Alternately this transfer may be facilitated via the use of a network based switching system such as those employed by Verizon Business (ECR) or CISCO (ICM). [0044] At 320 the original IP based call exits the Primary Preamble System 300 and terminates on the ACD/PABX 400. As with telephone calls conveyed over the PSTN 310 IP based calls may be switched using common methods such as SIP (session initiation protocol) redirect or SIP re-invite as specified in the SIP RFC 3261 promulgated by the IETF Network Working Group [0045] At 400 the telephone calls are processed by the ACD or PABX in which the determination of the best suited and available agent is made. One of ordinary skill in the art of contact centers will recognize the use of commonly available systems to achieve this agent selection and routing result. For example vendors such as Avaya Aspect and Alcatel manufacture telephone systems that can route calls to agents who are either local or remote using both PSTN and/or IP based facilities. [0046] Enroller Aspect. [0047] At 410 the original PSTN telephone call exits the ACD/PABX 400 and terminates on a telephone device 510. Alternately at 420 the original IP based telephone call terminates on a computer 550 with the use of a so called softphone or alternately on an IP based instrument 510. [0048] In addition, t he enroller aspect point 550 (agent workstation) may house an enrolment widget. In a preferred embodiment of the invention the enrolment widget will be implemented using a common computer language such as javascript, adobe flex or adobe AIR. Anyone of ordinary skill in the art of desktop software will be able to communicate with the biometric application gateway RESTIet 700 via HTTP. The purpose of the enrolment widget is to allow the enroller/agent 505 to verify the identity of the caller and match that verification to the biometric key of the caller. In addition, the widget may WO 2012/096901 PCT/US2012/020679 -8 be employed such that the biometric application gateway RESTIet 700 can send messages to the enroller agent 505 indicating the score of automated verification of a particular caller. [0049] Biometric Complex. [0050] The biometric complex is comprised of the biometric engine 600, the biometric application RESTIet 700 and the biometric application database 800. [0051] At 600 the biometric engine houses a highly specialized speech algorithm and scoring mechanism with which stored biometric keys (biometric verbal fingerprints) are matched in order to ascertain the identity of the caller. Such a biometric engine may be deployed using commonly available software. One of ordinary skill in the art may deploy biometric engine technology offered by vendors such as Nuance, Persay or Validvoice, for example. [0052] At 700 a software state controller for the biometric engine 600 and for the natural enrolment process is resident. This biometric application gateway may use a variety of computer applications and languages such as C++ or Java for example. [0053] In a preferred embodiment of the invention the Biometric Application Gateway 700 uses a secure RESTIet in order to process commands to and from the Biometric Engine 600 the Optional Preamble Engine 200; the Primary Preamble Server 300, the EnrolmentNerification Widget 550, and the Biometric Database 800. [0054] A variety of protocols may be used to communicate between the Biometric Application Gateway 700 and related components. For example the link between the Biometric Engine 60) and the Biometric Application Gateway 700 may be TCP/IP, HTTP or MRCP (media resource control protocol). For example, the protocol between Biometric Application Gateway 700 and the Optional Preamble Engine 200 and the Primary Preamble Server 300 m ay be HTTP, TCP/IP or MRCP. For example, the protocol connecting the Biometric Application Gateway 700 with the Enrolment/Verification Widget 550 may be HTTP, TCP/IP or some variety of web services. For example the Biometric Application Gateway (700) may communicate with the Biometric Application Database 800 using protocols such as JDBC or SQL Query. [0055] Depending on computer application language and protocol preference the biometric complex may be implemented as a monolithic application encapsulated in a single computer, or it may be implemented as separate logical systems, distributed over a network although acting as a whole.
WO 2012/096901 PCT/US2012/020679 -9 [0056] At 800 voiceprints and associated scoring data are stored in a biometric database which acts as a secure repository for voiceprints captured from Enrollee/Caller 105. Such stored biometric keys are used to compare subsequent utterances (ostensibly verbalized by a caller whom one wishes to verify) in order to ascertain a correct match between a stored biometric key versus a real time utterance from an Enrollee/Caller 105 who claims to be the individual who had previously and successfully been enrolled or verified. [0057] Conventional Automated Enrolment-Figure 2. [0058] A). Enrolment scenario in which no agent intervention is required. [0059] At 2000 the Conventional Automated Enrolment Process begins. At 2010 a caller who has yet to enroll calls the telephone number associated with the Interactive Voice Response system or ACD representing the target enterprise. At 2020 the telephone call is connected to the Optional Preamble Engine 200 or the Primary Preamble Server 300 depending on whether or not the linkage to the Biometric Complex 600, 700, 800 is achieved in a distributed fashion (with the Optional Preamble Engine sitting in front of the Primary Preamble Server) or incorporated as a function of the Primary Preamble Server. It is important to note that access to the Biometric Complex 600, 700, 800 may be achieved with a variety of methods including but not limited to the Optional Preamble Engine 200 being used as a passive listening device in front of the Primary Preamble Server 200 or as a shared resource in front of the Primary Preamble Server 300 or behind the Primary Preamble Server 300 as described in FIG. 1. This may also be achieved by providing access to the Biometric Complex 600, 700, 800 via non telephonic access methods (such as a computer microphone and VoIP arrangement) as described under the Enrollee Aspect 100 in FIG. 1. [0060] From a logic flow perspective step 2020, regardless of topology, serves the function of collecting the voiceprint of the caller so that it can be digitized and stored as a Biometric Key inside the Biometric Complex 600, 700, 800. [0061] At 2030 the Primary Preamble System 300 collects the standard Personal Identification Number (or customer or user number) (PIN) as typified by the majority of self-service systems currently deployed. This PIN number, which is used as a first stage identification may be spoken or entered via touch tone digits as will be appreciated by one of ordinary skill in the art of interactive voice response. [0062] In a preferred embodiment of the invention this PIN number is used to compare the customer record in a standard database to ascertain whether or not the calling party WO 2012/096901 PCT/US2012/020679 -10 has already enrolled. The step of collecting the PIN may optionally be skipped if it is possible to collect a biometric sample from a caller to verify the previous enrolment of the identity of the caller using a match to his stored biometric key. [0063] At 2040 the determination of whether or not the caller has previously enrolled is made. If the person is not enrolled "a not enrolled status flag" is stored in memory. This is then used to determine the appropriate call flow to be used by the IVR. If it is ascertained that the caller has already enrolled, the system executes a verification process as outlined in either FIGs. 4 or 5 or some other appropriate process. [0064] At 2050 the algorithms in the PABX or ACD processes the caller as appropriate to complete enrolment. Should the user already be enrolled then step 2060 is used to process a verification of the user. [0065] At 2070 it is determined whether the verification succeeded. If successful the caller is passed to the agent with the appropriate status. Optionally the caller may be passed to the agent with a fail status. These data are conveyed to the Agent Enrolment/Verification Widget 550 via the Biometric Application Gateway RESTIet 700 as described in FIG. 1. [0066] At 2080 the IVR asks the first enrolment question and collects the user response. [0067] At 2090 a determination is made whether sufficient voice data has been collected to complete an enrolment. If sufficient data have been collected then the data will be updated in the Biometric Database and the call optionally passed to an Agent at 2130 should a further transaction be required by the User. Alternately, control could be passed back to the top level IVR menu to allow the user to select a transaction. Alternately the call could be terminated if the only requirement was to complete enrolment. [0068] If at 2090 it is determined that insufficient enrolment voice data have b een captured a subsequent question will be asked at 2100. [0069] At 2110 a determination is made whether sufficient voice data have been collected to complete an enrolment. If sufficient voice data have been collected then the data will be updated in the Biometric Database and the call optionally passed to an Agent at 2130 should a further transaction be required by the User. Alternately control could be passed back to the top level IVR menu to allow the user to select a transaction. Alternately the call could be terminated if the only requirement was to complete enrolment.
WO 2012/096901 PCT/US2012/020679 -11 [0070] If at 2110 it is determined that insufficient enrolment voice data have b een captured a subsequent question will be asked at 2120. [0071] If sufficient voice data have been collected then the data will be updated in the Biometric Database and the call optionally passed to an Agent at 2130 should a further transaction be required by the User. Alternately control could be passed back to the top level IVR menu to allow the user to select a transaction. Alternately the call could be terminated if the only requirement was to complete enrolment. [0072] [0074] Should the enrolment fail appropriate information will be passed to the agent. These data are conveyed to the Agent Enrolment/Verification Widget 550 via the Biometric Application Gateway RESTIet 700 as described in FIG. 1. The agent may then choose to assist the user/enrollee perhaps with a process similar to that outlined in FIG. 3. [0073] Should the verification fail the caller will be passed to an agent at the same time any data that were stored relative to the caller having already been enrolled or scores of a current utterance will be forwarded to the appropriate Agent Enrollment/Verification Widget 550. [0074] In addition, the Primary Preamble System 300 will have forwarded the agent identity to the Biometric Application Gateway RESTIet 700 by extracting same via common CTI data path from the ACD/PABX 400. [0075] Alternately, the Optional Preamble Engine 200 may have a CTI data link to the ACD in order to extract chosen agent information. Alternately, the Biometric Application Gateway RESTIet 700 may extract the chosen agent identity information via common CTI links. It is important to note that there are a variety of methods for extracting from an ACD or PABX the chosen agent information. In addition, to the examples above it is also possible to extract the same information from a real time database. For example, the Biometric Database 800 as described in FIG. 1 could be used as a shared repository where the ACD, PABX, or CTI application gateway could deposit chosen agent data in real time. Such data may be extracted using a third party data proxy such as those supplied by Oracle Corporation. In addition, real time agent selection data may also be transmitted from a CTI application gateway to the Biometric Application Gateway RESTIet 700 via real time HTTP/REST or web service commands. [0076] Alternately, when the call gets to the agent desktop upon answer supervision or a key press from the agent or an uttered password from the agent the Biometric Complex 600, 700, 800 could be used to analyze the agent's voice in order for the system to WO 2012/096901 PCT/US2012/020679 -12 identify who the agent is who answered the phone to obviate completely the need for CTI data. In this scenario all of the enterprise agent's will have enrolled themselves in order to record a Biometric Key in the Biometric Complex 600, 700, 800. One of ordinary skill in the art will understand that this Biometric Key will be associated in the Biometric Database 800 with agent name, employee number, extension number and other data to establish an agent record in the database. In this fashion an inline verification of who the agent is may be achieved without the use of CTI and without the use of extracting any data from the legacy IVR or ACD currently installed. [0077] At 2140 the agent answers the phone and the Agent EnrolmentNerification Widget 550 is synchronized. The synchronization occurs either by the CTI method described above or the agent themselves biometrically verifying so that the Biometric Application Gateway RESTIet 700 knows to which Biometric Enrolment/Verification Widget 550 to send the caller data. [0078] In a preferred embodiment of the invention the Optional Preamble Engine 200 or one of its equivalents described above is used to idle the transmit timeslot on the telephone connection so that the caller does not hear what the agent needs to say in order for the agent to complete biometric verification. While the agent is verifying the Optional Preamble Engine 200 may be equipped with sound files to play out to the caller such as call progress tones and in this particular instance call ring back. Once the agent biometric routing has been accomplished the ring back tone will cease and both parties will be able to hear each other. [0079] At 2170 if the process failed the agent will be able to use the standard process established by that enterprise of resetting the PIN as described in step 2030. This is to ensure that a fraudulent caller who may have an intention to steal the identity of the true caller will not be able to use the PIN that gave him access at the beginning of the process 2000. [0080] Agent Assisted Natural Enrolment Process-Figure 3. [0081] At 3000 the Agent Assisted Natural Enrolment Process begins. At 3010 a caller who has yet to enroll calls the telephone number associated with the Interactive Voice Response system or ACD representing the target enterprise. At 3020 the telephone call is connected to the Optional Preamble Engine 200 or the Primary Preamble Server 300 depending on whether or not the linkage to the Biometric Complex 600, 700, 800 is achieved in a distributed fashion (with the Optional Preamble Engine sitting in front of the Primary Preamble Server) or incorporated as a function of the Primary Preamble Server.
WO 2012/096901 PCT/US2012/020679 -13 It is important to note that access to the Biometric Complex 600, 700, 800 may be achieved with a variety of methods including but not limited to the Optional Preamble Engine 200 being used as a passive listening device in front of the Primary Preamble Server 200 or as a shared resource in front of the Primary Preamble Server 300 or behind the Primary Preamble Server 300 as described in FIG. 1. This may also be achieved by providing access to the Biometric Complex 600, 700, 800 via non telephonic access methods (such as a computer microphone and VoIP arrangement) as described under the Enrollee Aspect 100 in FIG. 1. [0082] From a logic flow perspective step 3020, regardless of topology, serves the function of collecting the voiceprint of the caller so that it can be digitized and stored as a Biometric Key inside the Biometric Complex 600, 700, 800. [0083] At 3030 the Primary Preamble System 300 collects the standard Personal Identification Number (or customer or user number) (PIN) as typified by the majority of self-service systems currently deployed. This PIN number, which is used as a first stage identification may be spoken or entered via touch tone digits as is commonly achieved by any one of ordinary skill in the art of interactive voice response. [0084] In a preferred embodiment of the invention this PIN number is used to compare the customer record in a standard database to ascertain whether or not the calling party has already enrolled. The step of collecting the PIN may be optionally skipped if it is possible to collect a biometric sample from a caller to ascertain the previous enrolment of the identity of the caller using a match to his stored biometric key. [0085] At 3040 the determination of whether or not the caller has previously enrolled is made. If the person is not enrolled "a not enrolled status flag" is stored in memory and will be used later in the sequence to push that flag to the appropriate agent's Agent Enrolment/Verification Widget 550 in order to alert that agent that full Agent Assisted Natural Enrolment will be required. If it is ascertained that the caller has enrolled, the system will store the confidence score of the current utterances in order to alert the agent via the desktop widget whether subsequent live verification is appropriate. Such utterances will be solicited as part of a regular IVR process either by the Optional Preamble Engine 200 or the Primary Preamble System 300; regardless the Optional Preamble Engine 200 function will be listening and forwarding the recorded speech data to the Biometric Complex 600, 700, 800. [0086] At 3050 the algorithms in the PABX or ACD will route the caller to the appropriate agent at the same time any data that was stored relative to the caller having already been WO 2012/096901 PCT/US2012/020679 -14 enrolled or scores of a current utterance will be forwarded to the appropriate Agent EnrollmentNerification Widget 550. These data are conveyed to the Agent Enrolment/Verification Widget 550 via the Biometric Application Gateway RESTIet 700 as described in FIG. 1. [0087] In addition, the Primary Preamble System 300 will have forwarded the agent identity to the Biometric Application Gateway RESTIet 700 by extracting same via common CTI data path from the ACD/PABX 400. [0088] [0091] Alternately, the Optional Preamble Engine 200 may have a CTI data link to the ACD in order to extract chosen agent information. Alternately the Biometric Application Gateway RESTIet 700 may extract the chosen agent identity information via common CTI links. It is important to note that there are a variety of methods for extracting from an ACD or PABX the chosen agent information, In addition to the examples above, it is also possible to extract the same information from a real time database. For example, the Biometric Database 800 as described in FIG. 1 could be used as a shared repository where the ACD, PABX, or CTI application gateway could deposit chosen agent data in real time. Such data may be extracted using a third party data proxy such as those supplied by Oracle Corporation. In addition, real time agent selection data may also be transmitted from a CTI application gateway to the Biometric Application Gateway RESTIet 700 via real time HTTP/REST or web service commands. [0089] Alternately, when the call gets to the agent desktop upon answer supervision or a key press from the agent or an uttered password from the agent the Biometric Complex 600, 700, 800 could be used to analyze the agent's voice in order for the system to identify who the agent is who answered the phone to obviate completely the need for CTI data. In this scenario all of the enterprise agent's will have enrolled themselves in order to record Biometric Keys in the Biometric Complex 600, 700, 800. One of ordinary skill in the art will understand that this Biometric Key will be associated in the Biometric Database 800 with agent name, employee number, extension number and other data to establish an agent record in the database. In this fashion an inline verification of who the agent is may be achieved without the use of CTI and without the use of extracting any data from the legacy IVR or ACD currently installed. [0090] At 3060 the agent answers the phone and the Agent Enrolment/Verification Widget 550 is synchronized. The synchronization occurs either by the CTI method described above or the agent themselves being biometrically verified so that the Biometric WO 2012/096901 PCT/US2012/020679 -15 Application Gateway RESTlet 700 knows to which Biometric Enrolment/Verification Widget (550) to send the caller data. [0091] In a preferred embodiment of the invention the Optional Preamble Engine 200 or one of its equivalents described above shall be used to idle the transmit timeslot on the telephone connection so that the caller does not hear what the agent needs to say in order for the agent to perform biometric verification. While the agent is verifying, the Optional Preamble Engine 200 may be equipped with sound files to play out to the caller such as call progress tones and in this particular instance a call ring back tone. Once the agent biometric routing has been accomplished the ring back tone will cease and both parties will be able to hear each other. [0092] At 3070, once the EnrolmentNerification widget 550 has been synchronized and the call data and biometric scoring presented to the agent, the agent is prepared to begin asking verification questions, the caller's answers to which may be used to compare the caller biometric key with the stored one (assuming the caller has already enrolled), or in the case of an already enrolled caller whose biometric score is low the agent will be prepared to capture additional samples to bolster the confidence score of the biometric engine. [0093] In addition, in the case of the caller not having already enrolled the first verification challenge question of the caller may be asked. The Agent EnrolmentNerification Widget 550 may be equipped with a button that allows the agents to mute themselves after asking the questions so the Biometric Complex 600, 700, 800 is recording only voice print samples from a caller. [0094] At 3080 the agent makes a determination of the scoring at that point in time. [0095] At 3090 the agent will ask an additional verification question if in the normal course of live verification another question is required based on the business rules of that enterprise or if more voiceprint data are required to complete the enrolment. [0096] At 3100 the agent makes a determination of the scoring at that point in time. [0097] At 3110 the agent will ask an additional verification question if in the normal course of live verification another question is required based on the business rules of that enterprise or if more voiceprint data is required to complete the enrolment. [0098] At 3120 the agent is able to make a judgment based on the answers to the verification steps as to whether or not the caller is who they claim to be. The agent is then able to select a pass or fail button on the Agent Verification/Enrolment Widget 550 which WO 2012/096901 PCT/US2012/020679 -16 subsequently communicates with the Biometric Application Gateway RESTIet 700 that the caller has passed or failed. [0099] At 3130 the Biometric Application Gateway RESTIet 700 accesses the Biometric Database 800 in order to set the EnrolmentNerification pass/fail flag in the caller record. [00100] At 3140 if the process failed, the agent will be able to use the standard process established by that enterprise of resetting the PIN as described in step 3030. This is to ensure that a fraudulent caller who may have an intention to steal the identity of the true caller will not be able to use the PIN that gave him access at the beginning of the process 3000. [00101] Conventional Automated verification process-Fiqure 4. [00102] Conventional verification with no live agent intervention. [00103] At 4000 the Conventional Automated Verification Process begins. At 4010 a caller who has already enrolled calls the telephone number associated with the Interactive Voice Response system or ACD representing the target enterprise. At 4020 the telephone call is connected to the Optional Preamble Engine 200 or the Primary Preamble Server 300 depending on whether or not the linkage to the Biometric Complex 600, 700, 800 is achieved in a distributed fashion (with the Optional Preamble Engine sitting in front of the Primary Preamble Server) or incorporated as a function of the Primary Preamble Server. It is important to note that access to the Biometric Complex 600, 700, 800 may be achieved with a variety of methods including but not limited to the Optional Preamble Engine 200 being used as a passive listening device in front of the Primary Preamble Server 200 or as a shared resource in front of the Primary Preamble Server 300 or behind the Primary Preamble Server 300 as described in FIG. 1. This may also be achieved by providing access to the Biometric Complex 600, 700, 800 via non telephonic access methods (such as a computer microphone and VolP arrangement) as described under the Enrollee Aspect 100 in FIG. 1. [00104] From a logic flow perspective step 4020, regardless of topology, serves the function of collecting the voiceprint of the caller so that it can be digitized and stored as a Biometric Key inside the Biometric Complex 600, 700, 800. [00105] At 4030 the Primary Preamble System 300 collects the standard Personal Identification Number (or customer or user number) (PIN) as typified by the majority of self-service systems currently deployed. This PIN number, which is used as a first stage WO 2012/096901 PCT/US2012/020679 -17 identification may be spoken or entered via touch tone digits as is commonly achieved by one of ordinary skill in the art of interactive voice response. [001061 In a preferred embodiment of the invention this PIN number will be used to compare the customer record in a standard database to ascertain whether or not the calling party has already enrolled. [00107] Depending on the exact embodiment the step of collecting the PIN may be skipped if it is possible to collect a biometric sample from a caller to ascertain the previous enrolment of the identity of the caller using a match to their stored biometric key. [00108] At 4040 the determination of whether or not the caller has previously enrolled is made. If the person is not enrolled "a not enrolled status flag" is stored in memory. This is then used to determine the appropriate call flow to be used by the IVR. If it is ascertained that the caller has not enrolled, the system executes an Enrolment process as outlined in either FIG. 2 or 3 or some other appropriate process. [00109] At 4050 the call flows in the PABX or ACD will process the caller as appropriate to complete verification. [00110] At 4070 the IVR asks the first verification question and collects the user response. [00111] At 4080 a determination is made whether sufficient voice data have been collected to complete verification. If sufficient data have been collected then the data will be updated in the Biometric Database and the call optionally passed to an Agent at 4130 should a further transaction be required by the User. Alternately, control could be passed back to the top level IVR menu to allow the user to select a transaction. [00112] If at 4080 it is determined that insufficient verification voice data have been captured a subsequent question will be asked at 4090. [00113] At 4100 a determination is made whether sufficient voice data have been collected to complete verification. If sufficient data have been collected then the data will be updated in the Biometric Database and the call optionally passed to an Agent at 4120 should a further transaction be required by the User. Alternately control could be passed back to the top level IVR menu to allow the user to select a transaction. [00114] If at 4100 it is determined that insufficient verification voice data have been captured a subsequent question will be asked at 4110.
WO 2012/096901 PCT/US2012/020679 -18 [00115] If sufficient data have been collected then the data will be updated in the Biometric Database and the call optionally passed to an Agent at 4130 should a further transaction be required by the User. Alternately, control could be passed back to the top level IVR menu to allow the user to select a transaction. [00116] Should the verification fail appropriate information will be passed to the agent. These data are conveyed to the Agent Enrolment/Verification Widget 550 via the Biometric Application Gateway RESTIet 700 as described in FIG 1. The agent may then choose to assist the user/enrollee perhaps with a process similar to that outlined in FIG. 3. [00117] Should the verification fail the caller will be passed to an agent at the same time any data that was stored relative to the caller having already been enrolled or scores of a current utterance will be forwarded to the appropriate Agent Enrollment/Verification Widget 550. [00118] In addition the Primary Preamble System 300 will have forwarded the agent identity to the Biometric Application Gateway RESTIet 700 by extracting same via common CTI data path from the ACD/PABX 40). [00119] Alternately, the Optional Preamble Engine 200 may have a CTI data link to the ACD in order to extract chosen agent information. Alternately the Biometric Application Gateway RESTIet 700 may extract the chosen agent identity information via common CTI links. It is important to note that there are a variety of methods for extracting from an ACD or PABX the chosen agent information. In addition to the examples above it is also possible to extract the same information from a real time database. For example, the Biometric Database 800 as described in FIG. 1 could be used as a shared repository where the ACD, PABX, or CTI application gateway could deposit chosen agent data in real time. Such data may be extracted using a third party data proxy such as those supplied by Oracle Corporation. In addition, real time agent selection data may also be transmitted from a CTI application gateway to the Biometric Application Gateway RESTIet 700 via real time HTTP/REST or web service commands. [00120] Alternately, when the call gets to the agent desktop upon answer supervision or a key press from the agent or an uttered password from the agent the Biometric Complex 600, 700,800 could be used to analyze the agent's voice in order for the system to identify the agent who answered the phone to obviate completely the need for CTI data. In this scenario all of the enterprise agent's will have enrolled themselves in order to record a Biometric Key in the Biometric Complex 600, 700, 800. One of ordinary skill in the art will understand that this Biometric Key will be associated in the Biometric WO 2012/096901 PCT/US2012/020679 -19 Database 800 with agent name, employee number, extension number and other data to establish an agent record in the database. In this fashion an inline verification of the agent's identity can be achieved without the use of CTI and without the use of extracting any data from the legacy IVR or ACD currently installed. [00121] At 4120 the agent answers the phone and the Agent Enrolment/Verification Widget 550 is synchronized. The synchronization occurs either by the CTI method described above or by the agents themselves being biometrically verified so that the Biometric Application Gateway RESTIet 700 knows to which Biometric Enrolment/ Verification Widget 550 to send the caller data. [00122] In a preferred embodiment of the invention the Optional Preamble Engine 200 or one of its equivalents described above is used to idle the transmit timeslot on the telephone connection so that the caller does not hear what the agent needs to say in order for the agent to complete biometric verification. While the agent is verifying, the Optional Preamble Engine 200 may be equipped with sound files to play to the caller such as call progress tones and in this particular instance a ring back tone. Once the agent biometric routing has been accomplished the ring back tone will be cease and the parties will be able to hear each other. [00123] At 4140 if the process failed the agent will be able to use the standard process established by that enterprise of resetting the PIN as described in step 4030. This is to ensure that a fraudulent caller who may have an intention to steal the identity of the true caller will not be able to use the PIN that gave him access at the beginning of the process 4000. [00124] Agent assisted verification process-Figure 5. [00125] Live agent assisted verification. [00126] At 5000 the Live Agent Assisted Verification Process begins. At 3010 a caller who has already Enrolled calls the telephone number associated with the Interactive Voice Response system or ACD representing the target enterprise. At 3020 the telephone call is connected to the Optional Preamble Engine 200 or the Primary Preamble Server 300 depending on whether or not the linkage to the Biometric Complex 600, 700, 800 is achieved in a distributed fashion (with the Optional Preamble Engine sitting in front of the Primary Preamble Server) or incorporated as a function of the Primary Preamble Server. It is important to note that access to the Biometric Complex 600, 700, 800 may be achieved by means of a variety of methods including but not limited to the Optional WO 2012/096901 PCT/US2012/020679 -20 Preamble Engine 200 being used as a passive listening device in front of the Primary Preamble Server 200 or as a shared resource in front of the Primary Preamble Server 300 or behind the Primary Preamble Server 300 as described in FIG. 1. This may also be achieved by providing access to the Biometric Complex 600, 700, 800 via non telephonic access methods (such as a computer microphone and VoIP arrangement) as described under the Enrollee Aspect 100 in FIG. 1. [00127][From a logic flow perspective step 5020, regardless of topology, serves the function of collecting the voiceprint of the caller so that it can be digitized and stored as a Biometric Key inside the Biometric Complex 600, 700, 800. [00128] At 5030 the Primary Preamble System 300 collects the standard Personal Identification Number (or customer or user number) (PIN) as typified by the majority of self-service systems currently deployed. This PIN number, which is used as a first stage identification may be spoken or entered via touch tone digits as is apparent to any one of ordinary skill in the art of interactive voice response. [00129] In a preferred embodiment of the invention this PIN number will be used to compare the customer record in a standard database to ascertain whether or not the calling party has already enrolled. [00130] The step of collecting the PIN may be optionally skipped if it is possible to collect a biometric sample from a caller to ascertain the previous enrolment of the identity of the caller using a match to their stored biometric key. [00131] At 5040 the determination of whether or not the caller has previously enrolled is made. If the person is not enrolled "a not enrolled status flag" is stored in memory and at 5045 control is passed to the process as outlined in FIG. 3 from step 3050 where the flag will be used later in the sequence to push that flag to the appropriate agent's Agent Enrolment/Verification Widget 550 in order to alert that agent that full Agent Assisted Natural Enrolment will be required. If it is ascertained that the caller has already enrolled, the system will store the confidence score of the current utterances in order to alert the agent via the desktop widget whether subsequent live verification is appropriate. Such utterances will be solicited as part of a regular IVR process either by the Optional Preamble Engine 200 or the Primary Preamble System 300; in either case the Optional Preamble Engine 200 function will be listening and forwarding the recorded speech data to the Biometric Complex 600, 700, 800.
WO 2012/096901 PCT/US2012/020679 -21 [00 132] At 5050 the algorithms in the PABX or ACD will route the caller to the appropriate agent at the same time any data that was stored relative to the caller having already been enrolled or scores of a current utterance will be forwarded to the appropriate Agent EnrollmentVerification Widget 550. These data are conveyed to the Agent Enrolment/Verification Widget 550 via the Biometric Application Gateway RESTIet 700 as described in FIG. 1. [00133] In addition, the Primary Preamble System 300 will have forwarded the agent identity to the Biometric Application Gateway RESTIet 700 by extracting same via common CTI data path from the ACD/PABX 400. [00134] Alternately, the Optional Preamble Engine 200 may have a CTI data link to the ACD in order to extract chosen agent information. Alternatively, the Biometric Application Gateway RESTIet 700 may extract the chosen agent identity information via common CTI links. It is important to note that there are a variety of methods for extracting from an ACD or PABX the chosen agent information. In addition to the examples above it is also possible to extract the same information from a real time database. For example the Biometric Database 800 as described in FIG. 1 could be used as a shared repository where the ACD, PABX, or CTI application gateway could deposit chosen agent data in real time. Such data may be extracted using a third party data proxy such as those supplied by Oracle Corporation. In addition, real time agent selection data may also be transmitted from a CTI application gateway to the Biometric Application Gateway RESTIet 700 via real time HTTP/REST or web service commands. [00135] Alternately, when the call gets to the agent desktop upon answer supervision or a key press from the agent or an uttered password from the agent, the Biometric Complex 600, 700, 800 could be used to analyze the agent's voice in order for the system to identify the agent who answered the phone to obviate completely the need for CTI data. In this scenario all of the enterprise agent's will have enrolled themselves in order to record a Biometric Key in the Biometric Complex 600, 700, 800. It will be appreciated that this Biometric Key will be associated in the Biometric Database 800 with agent name, employee number, extension number and other data to establish an agent record in the database. In this fashion an inline verification of agent identity can be achieved without the use of CTI and without the use of extracting any data from the legacy IVR or ACD currently installed. [00136] At 5060 the agent answers the phone and the Agent Enrolment/Verification Widget 550 is synchronized. The synchronization occurs either by the CTI method WO 2012/096901 PCT/US2012/020679 -22 described above or the agents biometrically verifying themselves so that the Biometric Application Gateway RESTIet 700 knows to which Biometric EnrolmentNerification Widget 550 to send the caller data. [00137] In a preferred embodiment of the invention the Optional Preamble Engine 200 or one of its equivalents described above shall be used to idle the transmit timeslot on the telephone connection so that the caller does not hear what the agent needs to say in order for the agent to achieve biometric verification. While the agent is verifying, the Optional Preamble Engine (200) may be equipped with sound files to play to the caller such as call progress tones and in this particular instance ring back tones. Once the agent biometric routing has been accomplished the ring back tone will cease and the parties will be able to hear each other. [00138] At 5070, once the EnrolmentNerification widget 550 has been synchronized and the call data and biometric scoring presented to the agent, the agent is prepared to begin asking verification questions, the caller's answers to which may be used to compare the caller biometric key with the stored one (assuming the caller has already enrolled), or in the case of an already enrolled caller whose biometric score is low the agent will be prepared to capture additional samples to bolster the confidence score of the biometric engine. [00139] In addition, in the case of the caller not having already enrolled the first verification challenge question of the caller may be asked. The Agent Enrolment/Verification widget 550 may be equipped with a button that allows the agent to mute them self after asking the question so the Biometric Complex 600, 700, 800 is only recording voice print samples from a caller. [00140] At 5080 the agent makes a determination of the scoring at that point in time. [00141] At 5090 the agent will ask another verification question if in the normal course of live verification another question is required based on the business rules of that enterprise or if more voiceprint data are required to complete the enrolment. [00142] At 5100 the agent makes a determination of the scoring at that point in time. [00143] At 5110 the agent will ask another verification question if in the normal course of live verification another question is required based on the business rules of that enterprise or if more voiceprint data are required to complete the enrolment. [00144] At 5120 the agent is able to make a judgment based on the answers to the verification steps as to whether or not the caller is who they claim to be. The agent is then WO 2012/096901 PCT/US2012/020679 -23 able to select a pass or fail button on the Agent Verification/Enrolment Widget 550 which subsequently communicates with the Biometric Application Gateway RESTIet 700 that the caller has passed or failed. [00145] At 5130 the Biometric Application Gateway RESTIet 700 accesses the Biometric Database 800 in order to set the EnrolmentNerification pass/fail flag in the caller record. [00146] At 5140 if the process failed, the agent will be able to use the standard process established by that enterprise of resetting the PIN as described in step 5030. This is to ensure that a fraudulent caller who may have an intention to steal the identity of the true caller will not be able to use the PIN that gave him access at the beginning of the process 5000. [00147] The following claims are thus to be understood to include what is specifically illustrated and described above, what is conceptually equivalent, what can be obviously substituted and also what essentially incorporates the essential idea of the invention. Those skilled in the art will appreciate that various adaptations and modifications of the just-described preferred embodiment can be configured without departing from the scope of the invention. The illustrated embodiment has been set forth only for the purposes of example and that should not be taken as limiting the invention. Therefore, it is to be understood that, within the scope of the appended claims, the invention may be practiced other than as specifically described herein.

Claims (27)

1. A biometric voice identification system that can be used as an overlay on top of existing interactive voice response systems or a telephone routing 5 system to augment the interactive voice response system or telephone routing system with an enrolment and verification capability without making changes to the legacy interactive voice response system or telephone routing system, the system comprising: an enrollee aspect for accepting user audio input by means of PSTN, 10 SIP or VolP; an optional preamble engine and or a primary preamble system of deriving speech samples from user audio input; a biometric complex for processing- the speech samples, the biometric complex comprising: 15 a biometric engine for accepting the speech samples; a biometric application gateway for producing biometric keys from speech samples and comparing biometric keys; and a biometric database for storing biometric keys produced by the biometric application gateway; 20 an enrollment subsystem which uses input from a user to determine authenticity of a users identity whereupon if the user is determined to be authentic, biometric keys from the user's audio input are stored in the biometric database; and a verification subsystem which uses audio input from a user to 25 determine if the user matches a stored biometric key whereupon the user is verified. SUBSTITUTE SHEET L~TD 01 teMihf UIT D 2A\J T~eOCI,07A VVJ IT"YT 7TA7 TT?~ IPEA/KR 22 Nov. 2012 25
2. The biometric voice identification system according to claim 1 further comprising an enroller/agent aspect comprising: a live agent; means for transmitting the user's audio input to the live agent and for 5 transmitting the live agent's audio response to the user; an enrollment widget which receives and displays results from the biometric application gateway and which receives instruction from the live agent and implements the instructions.
3. The biometric voice identification system according to claim 2, 10 wherein the live agent makes a pass/fail judgment on a pending enrolment.
4. The biometric voice identification system according to claim 3, wherein said judgment is based on the user's answer to enrolment questions.
5. The biometric voice identification system according to claim 2, wherein the live agent makes a pass/fail judgment on a pending verification. 15
6. The biometric voice identification according to claim 5, wherein said judgment is based on output from the biometric application gateway.
7. The biometric voice identification system according to claim 5, wherein said judgment is based on the user's answer to verification questions.
8. The biometric voice identification according to claim 5, wherein 20 said judgment is based on output from the biometric application gateway and information provided by the user. SUBSTITUTE SHEET OTAI~ MA~I~T 'it.r-1ai wrn L I L~CVlLVV IPEA/KR 22 Nov. 2012 26
9. A biometric listening apparatus and software that can be used as an overlay on top of existing interactive voice response systems or a telephone routing system to augment the interactive voice response system or telephone routing system with an enrolment capability without making changes to the legacy 5 interactive voice response system or telephone routing system comprising: an enrollee aspect for accepting user audio input by means of PSTN, SIP orVolP; a biometric complex for processing the speech samples, the biometric complex comprising: 10 a biometric engine for. accepting the speech samples; a biometric application gateway for producing biometric keys from speech samples and comparing biometric keys; and a biometric database for storing biometric keys produced by the biometric application gateway; 15 an enrollment subsystem which uses input from a user to determine authenticity of a user's identity whereupon if the user is determined to be authentic, biometric keys from the user's audio input are stored in the biometric database; a live agent/enroller including means for transmitting the user's audio 20 input to the live agent and for transmitting the live agent's audio response to the user; and a verification subsystem which uses audio input from a user to determine if the user matches a stored biometric key whereupon the user is verified.
10, The biometric listening apparatus and software as per claim 9, 25 with the further ability to extract exclusive caller/enroller voice print data without sampling the agent/enroller voice print data to ensure a quality biometric token sample for the caller. SUBSTITUTE SHEET -' TA I-. -W :ItfIF A i f t AA i IPEA/KR 22 Nov. 2012 27
11. The biometric listening apparatus and software as per claim 9, with the further ability to extract exclusive agent/enroller voice print data without sampling the client/enroller voice print data to ensure a quality biometric token sample for the agent. 5
12. The biometric listening apparatus and software as per claim 14, with the further ability to extract exclusive caller/enroller voice print data without sampling the agent/enroller voice print data to ensure a quality biometric token sample for the caller.
13. The biometric listening apparatus and software as per claim 14, 10 with the further ability to extract exclusive agent/enroller voice print data without sampling the client/enroller voice print data to ensure a quality biometric token sample for the agent.
14. A live agent-facing software and methodology for use by enterprise live agents that provides synchronization between the biometric enrolment 15 system (caller-facing) with live agent biometric verification in order to ascertain the location of the agent to serve the caller without the use of legacy computer telephony infrastructure software to identify the agent, such back-to-back biometric dual verification having a common software control to create a self-contained bi directional verification system. 20
15. The live agent-facing software and methodology as per claim 14, that allows the conveyance of biometric scoring information In real time to the agent relative to the current caller, such software allowing the agent to use human judgment in directing the verification system in accepting the current score, or to continue the verification process by asking more questions to get more voice print 25 samples to heighten the confidence of the score or to deny the confidence of the score. SUBSTITUTE SHEET IPEA/KR 22 Nov. 2012 28
16. (currently amended) The live agent-facing software and methodology as per claim 14, that allows the conveyance of biometric scoring information in real time to the agent relative to the current caller, such software allowing the agent to use human judgment in directing the verification system in 5 accepting the current voice print sample, or to continue the enrolment process by asking more questions to get more voice print samples to add to the biometric voice print corpus for that caller.
17. (currently amended) The live agent-facing software and methodology as per claim 14, that allows the live agent to send a allowed, denied, or 10 other attribute score directly to the biometric enrolment system in order to add a second level of security and live verification to the verification process.
18. The live agent-facing software and methodology as per claim 14, that allows the live agent to send a allowed, denied, or other attribute score directly to the biometric enrolment system in order to add a third level of security and 15 live verification to the enrolment process.
19. The live agent-facing software and methodology as per claim 14, that allows the live agent to send a PIN (Personal Identification Code) reset command to the biometric enrolment system in order to disallow subsequent calls from a fraudulent individual using the former PIN code, such system alternately 20 being able to interface with third party systems including but not limited to interactive voice response and web self-service systems to reset the PIN codes on those systems.
20. (currently amended) The biometric listening apparatus and software as per claim 9 that allows for the collection of caller voice print data and 25 data collected live from an enterprise live agent; the combination of both automated biometric data collected and live agent data collected being substantially more secure than strictly automated data collected. SUBSTITUTE SHEET TZ AU4Ilhffl 61UtCTILOT ') A I F ., I I t/04,U I C .1,U I V IPEA/KR 22 Nov. 2012 29
21. The biometric listening apparatus and software as per claim 9 applied to GPS systems.
22. The biometric listening apparatus and software as per claim 9 applied to speech recognition systems. 5
23. The biometric listening apparatus and software as per claim 9 applied to soft phones and VolP networks.
24. The biometric listening apparatus and software as per claim 9 applied to ATM systems.
25, The biometric listening apparatus and software as per claim 9 10 applied to set top boxes.
26. The biometric listening apparatus and software as per claim 9 applied to in vehicle security and control systems.
27. The biometric listening apparatus and software as per claim 9 applied to embedded security systems. SUBSTITUTE SHEET
AU2012205747A 2011-01-10 2012-01-09 Natural enrolment process for speaker recognition Active AU2012205747B9 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US201161431402P 2011-01-10 2011-01-10
US61/431,402 2011-01-10
PCT/US2012/020679 WO2012096901A2 (en) 2011-01-10 2012-01-09 Natural enrolment process for speaker recognition

Publications (3)

Publication Number Publication Date
AU2012205747A1 AU2012205747A1 (en) 2013-07-18
AU2012205747B2 AU2012205747B2 (en) 2015-01-15
AU2012205747B9 true AU2012205747B9 (en) 2015-02-12

Family

ID=46507628

Family Applications (1)

Application Number Title Priority Date Filing Date
AU2012205747A Active AU2012205747B9 (en) 2011-01-10 2012-01-09 Natural enrolment process for speaker recognition

Country Status (5)

Country Link
EP (1) EP2663977A4 (en)
KR (1) KR101626302B1 (en)
AU (1) AU2012205747B9 (en)
CA (1) CA2823577A1 (en)
WO (1) WO2012096901A2 (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2523577A (en) * 2014-02-28 2015-09-02 Nokia Technologies Oy Speaker recognition method
GB2545534B (en) * 2016-08-03 2019-11-06 Cirrus Logic Int Semiconductor Ltd Methods and apparatus for authentication in an electronic device
CN108074577A (en) * 2016-11-09 2018-05-25 上海诺悦智能科技有限公司 Identity identifying method and identity authorization system
CN109102810B (en) * 2017-06-21 2021-10-15 北京搜狗科技发展有限公司 Voiceprint recognition method and device
US10592649B2 (en) * 2017-08-09 2020-03-17 Nice Ltd. Authentication via a dynamic passphrase
CN114270301A (en) * 2019-08-27 2022-04-01 索尼集团公司 Agent providing system, agent providing method, and recording medium
WO2022072655A1 (en) * 2020-10-01 2022-04-07 Pindrop Security, Inc. Enrollment and authentication over a phone call in call centers

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
AU7304798A (en) * 1996-11-22 1998-06-10 T-Netix, Inc. Voice recognition for information system access and transaction process ing
US6078807A (en) 1997-08-26 2000-06-20 International Business Machines Corporation Telephony fraud detection using voice recognition techniques
US6092192A (en) * 1998-01-16 2000-07-18 International Business Machines Corporation Apparatus and methods for providing repetitive enrollment in a plurality of biometric recognition systems based on an initial enrollment
US7269737B2 (en) * 2001-09-21 2007-09-11 Pay By Touch Checking Resources, Inc. System and method for biometric authorization for financial transactions
US7054811B2 (en) * 2002-11-06 2006-05-30 Cellmax Systems Ltd. Method and system for verifying and enabling user access based on voice parameters
US7212613B2 (en) 2003-09-18 2007-05-01 International Business Machines Corporation System and method for telephonic voice authentication
US7254383B2 (en) * 2004-07-30 2007-08-07 At&T Knowledge Ventures, L.P. Voice over IP based biometric authentication
US20060106605A1 (en) * 2004-11-12 2006-05-18 Saunders Joseph M Biometric record management
JP4679919B2 (en) * 2005-02-10 2011-05-11 Kddi株式会社 Distributed speaker verification device, distributed speaker verification terminal device, and computer program
US20060277043A1 (en) * 2005-06-06 2006-12-07 Edward Tomes Voice authentication system and methods therefor
US7340042B2 (en) * 2005-10-21 2008-03-04 Voiceverified, Inc. System and method of subscription identity authentication utilizing multiple factors
WO2008074076A1 (en) 2006-12-19 2008-06-26 Torqx Pty Limited Confidence levels for speaker recognition
US8010367B2 (en) * 2006-12-22 2011-08-30 Nuance Communications, Inc. Spoken free-form passwords for light-weight speaker verification using standard speech recognition engines

Also Published As

Publication number Publication date
EP2663977A2 (en) 2013-11-20
KR20130133828A (en) 2013-12-09
KR101626302B1 (en) 2016-06-01
WO2012096901A3 (en) 2012-10-26
CA2823577A1 (en) 2012-07-19
EP2663977A4 (en) 2016-12-28
AU2012205747B2 (en) 2015-01-15
AU2012205747A1 (en) 2013-07-18
WO2012096901A2 (en) 2012-07-19

Similar Documents

Publication Publication Date Title
AU2012205747B9 (en) Natural enrolment process for speaker recognition
US11210461B2 (en) Real-time privacy filter
US7340042B2 (en) System and method of subscription identity authentication utilizing multiple factors
US9813551B2 (en) Multi-party conversation analyzer and logger
US9837079B2 (en) Methods and apparatus for identifying fraudulent callers
US8812319B2 (en) Dynamic pass phrase security system (DPSS)
US8885797B2 (en) Systems and methods for providing network-based voice authentication
US6327345B1 (en) Method and system therefor of establishing an acceptance threshold for controlling fraudulent telephone calls
US20120284026A1 (en) Speaker verification system
US11606461B2 (en) Method for training a spoofing detection model using biometric clustering
EP4002900A1 (en) Method and device for multi-factor authentication with voice based authentication

Legal Events

Date Code Title Description
SREP Specification republished
FGA Letters patent sealed or granted (standard patent)
PC Assignment registered

Owner name: VALIDVOICE, LLC

Free format text: FORMER OWNER(S): SAMUEL, GARFIELD; GARCIA, WILSON