(19)
(11)EP 3 001 343 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
04.11.2020 Bulletin 2020/45

(21)Application number: 15186516.9

(22)Date of filing:  23.09.2015
(51)International Patent Classification (IPC): 
G06F 21/32(2013.01)

(54)

SYSTEM AND METHOD OF ENHANCED IDENTITY RECOGNITION INCORPORATING RANDOM ACTIONS

SYSTEM UND VERFAHREN ZUR VERBESSERTEN IDENTITÄTSERKENNUNG DURCH INKORPORATION VON ZUFALLSHANDLUNGEN

SYSTÈME ET PROCÉDÉ DE RECONNAISSANCE D'IDENTITÉ AMÉLIORÉE INCORPORANT DES ACTIONS ALÉATOIRES


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 26.09.2014 US 201414497769

(43)Date of publication of application:
30.03.2016 Bulletin 2016/13

(73)Proprietor: Ademco Inc.
Golden Valley, MN 55422 (US)

(72)Inventors:
  • JELAVIC, Robert
    Morris Plains, NJ New Jersey 07950 (US)
  • OH, Eric
    Morris Plains, NJ New Jersey 07950 (US)

(74)Representative: Murgitroyd & Company 
Murgitroyd House 165-169 Scotland Street
Glasgow G5 8PL
Glasgow G5 8PL (GB)


(56)References cited: : 
US-A1- 2012 130 714
US-A1- 2013 227 651
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    FIELD



    [0001] The application pertains to authentication systems and methods usable by individuals. More particularly, the application pertains to regional monitoring, security or access systems which combine anatomical characteristics with responses to requested random behaviors.

    BACKGROUND



    [0002] There are many commercial applications that need to authenticate users. For example, door access systems such as found in hotels, businesses or the like. Monitoring systems or system control panels need to verify that the expected user is in fact the one interfacing with the system or control panel keypad.

    [0003] Patent document number US2013/227651A1 describes an approach for enabling multi-factor biometric authentication of a user of a mobile device. A biometric authenticator captures, via a mobile device, first and second biometric data for a user. The biometric authentication further associates the first biometric data and the second biometric data. The biometric authenticator then initiates a multi-factor authentication procedure that utilizes the first biometric data and the second biometric data to authenticate the user based on the association.

    [0004] Known facial and voice recognition systems are less than 100% accurate. Pre-recorded video/imagery/audio of the authorized user can be used to defeat these systems in security applications. Additionally, passwords for security systems or control panels can also be easily compromised if a different user learns the code.

    [0005] There is thus a continuing need to improve anatomically based authentication systems.

    [0006] The present invention in its various aspects is as set out in the appended claims.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0007] Fig. 1 illustrates a combined block and flow diagram of an embodiment hereof.

    DETAILED DESCRIPTION



    [0008] While embodiments can take many different forms, specific embodiments thereof are shown in the drawings and will be described herein in detail with the understanding that the present disclosure is to be considered as an exemplification of the principles hereof, as well as the best mode of practicing same. No limitation to the specific embodiment illustrated is intended.

    [0009] In one aspect, facial and/or voice recognition systems or circuitry can be combined with random actions requested by the authentication system to confirm both that the detected imagery/audio is from an authorized user and that the user is present in person. In a disclosed, exemplary system, the user, or individual seeking service, or access, would be prompted to perform one or more random actions which cannot be anticipated.

    [0010] The system would require the user, or requester, to perform the action(s) within a set period of time. Some examples of these actions include one or more of:

    touching your nose with your right thumb

    touching your left ear

    smiling

    making a sad face

    reading a random passage

    turning your head left



    [0011] User facial/voice training during setup and continuous adaption thereafter can be implemented by embodiments hereof to improve the performance of authenticating the requester, or, user. For example, voice training would require the authorized requester, or, user to read a script that characterizes their voice. Specifically to increase security, the training script need not include all of the words in the random passages. In this circumstance, the system would both authenticate the requester based on the correct sounds and apply the new words being introduced once the user was authenticated.

    [0012] In another embodiment the system could authenticate the user by whatever criterion was acceptable at the time. Then, it would ask the user to smile or exhibit another gesture or facial image, and learn the new facial biometrics for the action to be used at some later time. At some point later, the system could be updated or may rely on cloud based services to change the collection of actions used to randomize the authentication process.

    [0013] In the case of failed attempts, the owner on file can be notified along with a video clip of the failed attempt. As a backup for critical applications such as for a security panel, failed attempts can be immediately viewed and evaluated by a human operator at a central control station. As a deterrent to attempts at defeating the system, recorded video can also be used for criminal prosecution.

    [0014] In addition to security panels, this approach can be used to increase the security for any user identity application including but not limited to smartphone access, computers/laptops, security door access, timecards or banking ATMs. Websites could rely on the approach instead of passwords to login to sites. By requiring the user to perform a random action, there is a higher certainty in both that the user is present and of their identity.

    [0015] In yet another aspect, the system can instruct the user to perform random actions while applying a combination of voice and facial recognition algorithms relative to captured audio and video. Both embedded and Internet based (cloud based) systems are available.

    [0016] For voice recognition, the random phrases could be partially trained. So, some words or sounds were part of the initial training and then the phrases are broken down into structural units of speech (i.e. phonemes) so that the user's voice is evaluated against the training sample rather than the whole phrase. The phonemes that were not part of the original training could be used to refine the original training once all of the other authentication checks passed. The associations betweeen the phonemes and collected identity data are managed in the Authorized User Reference Database. For voice recognition, a Random Phrase Generator can create a new challenge phrase based on the identified phonemes and phonetic dictionaries. One such alphabet for phonemes is the International Phonetic Alphabet. This system could be used for any spoken language.

    [0017] For facial recognition, separate frames of video can be analyzed. In selected embodiments, the changes in key dimensions such as size/shape/distances/color of the eyes, eyebrows, chin, face outline, teeth or mouth relative to the nose or other facial structures can be evaluated. The identity of the requester, or user, can be evaluated through the reading of text, in conjunction with voice and through motion such as the way the hand may touch the face (or the shape of the hand itself).

    [0018] It will be understood that neither the type nor the form of use of a randomizing request are limitations hereof. All such randomizing events come within the scope hereof.

    [0019] Fig. 1 illustrates an integrated system 10 and flow diagram 100 illustrating an embodiment hereof. System 10 includes one or more user or requester input devices 12, 12a which can be the same or different. Representative input devices include key pads, touch screens, card readers, wireless receivers and the like without limitation.

    [0020] An unknown user or requester U1 can make an access request via input device 12, as at 102, while providing minimal identification information. Alternately, an access can be provided by user U2 in combination with partial identifying information, via input device and circuits 12a.

    [0021] The user inputs can be coupled to identification circuits 14 which can select a random identification action, as at 104. An updatable identification data base 16 can be interrogated by the circuits 14 which can select the random identification action.

    [0022] The selected identification action can be presented at 106, in a verbal or visual format via user output device and related circuits 20. Representative random action requests can include a text request displayed on a screen of the device 20. For example, read this ..., or do this... Alternately, audio instructions can be output via device 20, such as, repeat the following. In yet another alternate, a visual behavior such as touch your nose, or, touch your ear, can be requested.

    [0023] The provided responses are received at a response input device 22. Device 22 can include one or more microphones or cameras. As illustrated, at 108, the received audio or video characteristics of the user, or requester such as U1, or U2 can be recorded, analyzed and measured as at 108.

    [0024] The received responses, as may be processed including phonetic data, facial measurements, relative timing or pitch contours, all without limitation, can be stored in one or more storage devices as at 112. Evaluation circuits 28 coupled to the response input device 22 and an authorized user reference database 30 can compare the received responses to expected responses, as at 114. In this regard, where partial identifying information has been provided via device and circuits 12a, that information can be incorporated into the evaluation by circuits 28.

    [0025] Depending on the results of the comparison at 114, a determination can be made, as at 116 as to the identity of the user, and whether the user's request will be granted.

    [0026] Those of skill will understand that a variety of homophones and common phonetic sounds can be incorporated into information stored in the identification action data base 16. Some examples of these and their International Phonetic Alphabet translations include:

    bottle / model (match "al" from "batal / madal")

    flower / plow (match "aw" from "flawar / plaw")

    Here / hear / ear (match "Ir" from "hIr / hIr / Ir")

    Plate / eight / ate (match "et" from "plet / et / et")

    Greenery / scenery (match "inari" from "grinari / sinari")

    Explain / plane / airplane / plain (match "plen" from iksplen / plen / εrplen / plen)

    To / two / too (tu / tu / tu)

    Beat / beet (bit / bit)

    Air / heir (εr / εr)

    Hare / hair (hεr / hεr)

    Bare / Bear (bεr / bεr)

    Know / no (no / no)

    Wait / weight (wet / wet)

    Flowers in the open plain (match aw

    rz/ð

    /plen from "flaw

    rz an ða opan plen")

    I spent hours on the airplane (match aw

    rz/ð

    /plen from "aj spεnt awarz an ða εrplen")



    [0027] The random action selection process, as at 104, can request a user, or requester to perform a second identification action to collect new identity references for future access once the user has been authenticated. Alternately, the user can be asked to perform a combined action that authenticates while collecting additional identity information, such as reusing words or phrases or phonemes or using homophones. Advantageously, in both modes, the system performs continuous adaption to tune authentication based on known reference information to determine acceptable variability of received responses.

    [0028] In summary, the combination of these approaches with a randomized action determined by the secure system can increase the certainty of the identity of the user. In a further aspect, the above described method can also be used by a website to test for human interaction without identification. By requesting a user, or requester, to perform random actions as described herein it becomes more difficult for an intruding computer system to generate correct audio and video responses in an allotted time interval. The likelihood of successful intrusion is thereby reduced.

    [0029] A simplified version of this computing system could also be used as a "reverse Turing test" administered to simply determine whether the unknown user is a human by challenging them to perform these random actions. For example, a website could test the unknown user to perform the action and test for basic facial/vocal matches to prevent bots from posting spam or skewing online polls. This would be applicable for low security needs as well as where the bandwidth, processing and/or database constraints are limited.

    [0030] From the foregoing, it will be observed that numerous variations and modifications may be effected without departing from the scope of the appended claims. It is to be understood that no limitation with respect to the specific apparatus illustrated herein is intended or should be inferred. It is, of course, intended to cover by the appended claims all such modifications as fall within the scope of the claims.

    [0031] Further, logic flows depicted in the figures do not require the particular order shown, or sequential order, to achieve desirable results.


    Claims

    1. A method comprising:

    receiving initial biometric data related to an authorized user during a training phase, the initial biometric data comprising audio data of a voice of the authorized user when reading a training script comprising a plurality of phonemes;

    storing the initial biometric data in an authorized user database (30);

    receiving a request for service (102) including an identifier from a requesting user;

    evaluating the identifier and requesting a random physiological response (112) from the requesting user, the random physiological response (112) requested comprising the requesting user speaking a random word or phrase and the requesting user performing one action selected from a group consisting of the requesting user touching his or her nose, the requesting user touching his or her ear, and the requesting user turning his or her head;

    receiving the random physiological response from the requesting user;

    evaluating the random physiological response received;

    determining whether the random physiological response received corresponds to a predetermined response (114); and

    determining that the requesting user is the authorized user when the random physiological response received corresponds to the predetermined response.


     
    2. A system comprising:

    a user input device(12);

    training circuits coupled to the user input device (12), the training circuits configured to (1) receive initial biometric data related to an authorized user during a training phase, the initial biometric data comprising audio data of a voice of the authorized user when reading a training script comprising a plurality of phonemes, and (2) store the initial biometric data in an authorized user database;

    identification circuits (14) coupled to the user input device (12), the identification circuits configured to receive a request for service (102) including an identifier from a requesting user; select and generate a further request, for a random authentication behavior, the random authentication behavior requested being in the form of a random physiological response and comprising the requesting user speaking a random word or phrase and the requesting user performing one action selected from a group consisting of the requesting user touching his or her nose, the requesting user touching his or her ear, and the requesting user turning his or her head; and

    output circuits (20) configured to present the request for the random authentication behavior audibly or visually and further comprising evaluation circuits (28) configured to receive a user response to the request and detect whether the user response corresponds to an expected response; and configured to determine that the requesting user is the authorized user when the random physiological response received corresponds to the predetermined response.


     
    3. The system as in claim 2 further comprising a regional monitoring system including the identification circuits and the evaluation circuits.
     
    4. The system of claim 2 or claim 3 wherein the output circuits (20) are configured to present the request for the random-authentication behavior both audibly and visually.
     
    5. The system as in claim 2 wherein the user input device (12) includes at least one of a video camera directed to a viewing region and a microphone.
     
    6. The system as in claim 5 wherein at least one of the video camera and the microphone is coupled to the identification circuits (14).
     
    7. The system as in claim 2 wherein the evaluation circuits (28) are configured to store the expected response .
     
    8. The system as in claim 7 wherein the evaluation circuits (28) are configured to request new responses from the authorized user and store representations of the new responses for use in subsequent authentication processes.
     
    9. The system as in claim 8 further comprising correspondence processing circuits configured for frequency domain processing, mathematical processing, or phoneme-type processing.
     
    10. The system as in claim 8 wherein, responsive to a failure of correspondence between the user response and the expected response, the correspondence processing circuits are configured to forward video of a failed attempt to a remote monitoring station.
     


    Ansprüche

    1. Ein Verfahren, das Folgendes beinhaltet:

    Empfangen von anfänglichen biometrischen Daten, die im Zusammenhang mit einem autorisierten Benutzer stehen, während einer Trainingsphase, wobei die anfänglichen biometrischen Daten Audiodaten einer Stimme des autorisierten Benutzers während des Lesens eines Trainingsskripts, das eine Vielzahl von Phonemen beinhaltet, beinhalten;

    Speichern der anfänglichen biometrischen Daten in einer Datenbank (30) für autorisierte Benutzer;

    Empfangen einer Dienstanforderung (102), die eine Kennung von einem anfordernden Benutzer enthält;

    Auswerten der Kennung und Anfordern einer zufälligen physiologischen Reaktion (112) von dem anfordernden Benutzer, wobei die angeforderte zufällige physiologische Reaktion (112) beinhaltet, dass der anfordernde Benutzer ein zufälliges Wort oder einen zufälligen Ausdruck ausspricht und der anfordernde Benutzer eine Aktion ausführt, die aus einer Gruppe ausgewählt ist, die aus Folgendem besteht: der/die anfordernde Benutzer/in berührt seine oder ihre Nase, der/die anfordernde Benutzer/in berührt sein oder ihr Ohr und der/die anfordernde Benutzer/in dreht seinen oder ihren Kopf;

    Empfangen der zufälligen physiologischen Reaktion von dem anfordernden Benutzer; Auswerten der empfangenen zufälligen physiologischen Reaktion;

    Bestimmen, ob die empfangene zufällige physiologische Reaktion mit einer vorgegebenen Reaktion (114) übereinstimmt; und

    Bestimmen, dass der anfordernde Benutzer der autorisierte Benutzer ist, wenn die empfangene zufällige physiologische Reaktion mit der vorgegebenen Reaktion übereinstimmt.


     
    2. Ein System, das Folgendes beinhaltet:

    eine Benutzereingabevorrichtung (12);

    Trainingsschaltungen, die an die Benutzereingabevorrichtung (12) gekoppelt sind, wobei die Trainingsschaltungen konfiguriert sind, um (1) während einer Trainingsphase anfängliche biometrische Daten zu empfangen, die im Zusammenhang mit einem autorisierten Benutzer stehen, wobei die anfänglichen biometrischen Daten Audiodaten einer Stimme des autorisierten Benutzers während des Lesens eines Trainingsskripts, das eine Vielzahl von Phonemen beinhaltet, beinhalten, und (2) die anfänglichen biometrischen Daten in einer Datenbank für autorisierte Benutzer zu speichern;

    Identifikationsschaltungen (14), die an die Benutzereingabevorrichtung (12) gekoppelt sind, wobei die Identifikationsschaltungen für Folgendes konfiguriert sind: Empfangen einer Dienstanforderung (102), die eine Kennung von einem anfordernden Benutzer enthält; Auswählen und Generieren einer weiteren Anforderung für ein zufälliges Authentifizierungsverhalten, wobei das angeforderte zufällige

    Authentifizierungsverhalten die Form einer zufälligen physiologischen Reaktion aufweist und beinhaltet, dass der anfordernde Benutzer ein zufälliges Wort oder einen zufälligen Ausdruck ausspricht und der anfordernde Benutzer eine Aktion ausführt, die aus einer Gruppe ausgewählt ist, die aus Folgendem besteht: der/die anfordernde Benutzer/in berührt seine oder ihre Nase, der/die anfordernde Benutzer/in berührt sein oder ihr Ohr und der/die anfordernde Benutzer/in dreht seinen oder ihren Kopf; und

    Ausgabeschaltungen (20), die konfiguriert sind, um die Anforderung für das zufällige Authentifizierungsverhalten hörbar oder sichtbar zu präsentieren, und ferner Auswertungsschaltungen (28) beinhalten, die konfiguriert sind, um eine Benutzerreaktion auf die Anforderung zu empfangen und um zu detektieren, ob die Benutzerreaktion mit einer erwarteten Reaktion übereinstimmt; und konfiguriert sind, um zu bestimmen, dass der anfordernde Benutzer der autorisierte Benutzer ist, wenn die empfangene zufällige physiologische Reaktion mit der vorgegebenen Reaktion übereinstimmt.


     
    3. System gemäß Anspruch 2, das ferner ein regionales Überwachungssystem beinhaltet, das die Identifikationsschaltungen und die Auswertungsschaltungen enthält.
     
    4. System gemäß Anspruch 2 oder Anspruch 3, wobei die Ausgabeschaltungen (20) konfiguriert sind, um die Anforderung für das zufällige Authentifizierungsverhalten sowohl hörbar als auch sichtbar zu präsentieren.
     
    5. System gemäß Anspruch 2, wobei die Benutzereingabevorrichtung (12) mindestens eines von einer Videokamera, die auf einen Sichtbereich ausgerichtet ist, und einem Mikrofon enthält.
     
    6. System gemäß Anspruch 5, wobei mindestens eines von der Videokamera und dem Mikrofon an die Identifikationsschaltungen gekoppelt ist (14).
     
    7. System gemäß Anspruch 2, wobei die Auswertungsschaltungen (28) konfiguriert sind, um die erwartete Reaktion zu speichern.
     
    8. System gemäß Anspruch 7, wobei die Auswertungsschaltungen (28) konfiguriert sind, um neue Reaktionen von dem autorisierten Benutzer anzufordern und Darstellungen der neuen Reaktionen zum Verwenden in nachfolgenden Authentifizierungsprozessen zu speichern.
     
    9. System gemäß Anspruch 8, das ferner Übereinstimmungsverarbeitungsschaltungen beinhaltet, die zur Frequenzdomänenverarbeitung, mathematischen Verarbeitung oder Phonemtypverarbeitung konfiguriert sind.
     
    10. System gemäß Anspruch 8, wobei, als Reaktion darauf, dass die Benutzerreaktion und die erwartete Reaktion nicht übereinstimmen, die Übereinstimmungsverarbeitungsschaltungen konfiguriert sind, um das Video eines fehlgeschlagenen Versuchs an eine entfernte Überwachungsstation weiterzuleiten.
     


    Revendications

    1. Un procédé comprenant :

    la réception de données biométriques initiales relatives à un utilisateur autorisé durant une phase d'entraînement, les données biométriques initiales comprenant des données audio d'une voix de l'utilisateur autorisé lisant un script d'entraînement comprenant une pluralité de phonèmes ;

    le stockage des données biométriques initiales dans une base de données d'utilisateur autorisé (30) ;

    la réception d'une requête de service (102) incluant un identifiant de la part d'un utilisateur requérant ;

    l'évaluation de l'identifiant et la requête d'une réponse physiologique aléatoire (112) de la part de l'utilisateur requérant, la réponse physiologique aléatoire (112) requise comprenant le fait que l'utilisateur requérant énonce un mot ou une phrase aléatoires et que l'utilisateur requérant réalise une action sélectionnée dans un groupe constitué du fait que l'utilisateur requérant se touche le nez, que l'utilisateur requérant se touche l'oreille, et que l'utilisateur requérant tourne la tête ;

    la réception de la réponse physiologique aléatoire de la part de l'utilisateur requérant ; l'évaluation de la réponse physiologique aléatoire reçue ;

    la détermination quant à savoir si la réponse physiologique aléatoire reçue correspond à une réponse préétablie (114) ; et

    la détermination que l'utilisateur requérant est l'utilisateur autorisé lorsque la réponse physiologique aléatoire reçue correspond à la réponse préétablie.


     
    2. Un système comprenant :

    un dispositif de saisie utilisateur (12) ;

    des circuits d'entraînement couplés au dispositif de saisie utilisateur (12), les circuits d'entraînement étant configurés pour (1) recevoir des données biométriques initiales relatives à un utilisateur autorisé durant une phase d'entraînement, les données biométriques initiales comprenant des données audio d'une voix de l'utilisateur autorisé lisant un script d'entraînement comprenant une pluralité de phonèmes, et (2) stocker les données biométriques initiales dans une base de données d'utilisateur autorisé ;

    des circuits d'identification (14) couplés au dispositif de saisie utilisateur (12), les circuits d'identification étant configurés pour recevoir une requête de service (102) incluant un identifiant de la part d'un utilisateur requérant ; pour sélectionner et générer une requête supplémentaire, pour un comportement d'authentification aléatoire, le comportement d'authentification aléatoire requis se présentant sous la forme d'une réponse physiologique aléatoire et comprenant le fait que l'utilisateur requérant énonce un mot ou une phrase aléatoires et que l'utilisateur requérant réalise une action sélectionnée dans un groupe constitué du fait que l'utilisateur requérant se touche le nez, que l'utilisateur requérant se touche l'oreille, et que l'utilisateur requérant tourne la tête ; et

    des circuits de sortie (20) configurés pour présenter la requête pour le comportement d'authentification aléatoire de façon sonore ou visuelle et comprenant en outre des circuits d'évaluation (28) configurés pour recevoir une réponse de l'utilisateur à la requête et détecter si la réponse de l'utilisateur correspond à une réponse attendue ; et

    configurés pour déterminer que l'utilisateur requérant est l'utilisateur autorisé lorsque la réponse physiologique aléatoire reçue correspond à la réponse préétablie.


     
    3. Le système tel que dans la revendication 2 comprenant en outre un système de surveillance régional incluant les circuits d'identification et les circuits d'évaluation.
     
    4. Le système de la revendication 2 ou de la revendication 3, dans lequel les circuits de sortie (20) sont configurés pour présenter la requête pour le comportement d'authentification aléatoire de façon aussi bien sonore que visuelle.
     
    5. Le système tel que dans la revendication 2 dans lequel le dispositif de saisie utilisateur (12) inclut au moins un élément parmi une caméra vidéo dirigée sur une région de visualisation et un microphone.
     
    6. Le système tel que dans la revendication 5 dans lequel au moins un élément parmi la caméra vidéo et le microphone est couplé aux circuits d'identification (14).
     
    7. Le système tel que dans la revendication 2 dans lequel les circuits d'évaluation (28) sont configurés pour stocker la réponse attendue.
     
    8. Le système tel que dans la revendication 7 dans lequel les circuits d'évaluation (28) sont configurés pour requérir de nouvelles réponses de la part de l'utilisateur autorisé et stocker des représentations des nouvelles réponses pour utilisation dans des processus d'authentification subséquents.
     
    9. Le système tel que dans la revendication 8 comprenant en outre des circuits de traitement de correspondence configurés pour un traitement de domaine de fréquence, un traitement mathématique, ou un traitement type phonèmes.
     
    10. Le système tel que dans la revendication 8 dans lequel, en réponse à un échec de correspondence entre la réponse de l'utilisateur et la réponse attendue, les circuits de traitement de correspondence sont configurés pour faire suivre la vidéo d'une tentative ayant échoué à une station de surveillance à distance.
     




    Drawing








    Cited references

    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description