DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Election/Restrictions
Applicant’s election without traverse of Invention II (2–7, 11–13, and 16–18) in the reply filed on April 20, 2026 is acknowledged. The application has pending claims 1-20 (withdrawn claims 8-9, 14 and 19-20 are withdrawn from further consideration).
Claim Rejections - 35 USC § 112(a)
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1–7, 11–13, and 16–18 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claims contain subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Independent claim 1 recites, in part, “determining, by the event identification circuitry, a signature parameter set for the signature” and “storing, by signature management circuitry, the signature and corresponding signature parameter set in a signature profile associated with the user within a signature corpus”. Independent claims 10 and 15 recite corresponding apparatus and computer-program-product limitations. Dependent claims 5, 13, and 18 further rely on the signature parameter set by reciting similarity scoring based on an inferred similarity between a corresponding signature parameter set for a stored signature and a signature parameter set for the candidate signature.
The specification describes the “signature parameter set” broadly and includes examples such as the physical condition of the signee, the emotional state of the signee, and the weather. However, the specification does NOT provide sufficient technical disclosure explaining how the claimed event identification circuitry determines such parameters for a signature. In particular, the specification does not disclose sufficient sensors, data sources, signal-processing steps, algorithms, model architecture, training data, feature-extraction techniques, or hardware configuration for determining a user’s emotional state or physical condition merely in connection with receiving or processing a signature. The claims are not limited to objective signature-capture parameters. Rather, in view of the specification, the claimed “signature parameter set” encompasses subjective and complex user-state and environmental variables, including emotional state, physical condition, and weather. The specification does not provide working examples showing how those parameters are measured, inferred, verified, normalized, or stored as part of the claimed signature parameter set.
The nature of the invention is computer-based signature management and signature verification, but the undisclosed subject matter includes determining subjective human states such as emotion and physical condition, which is not predictable from signature input alone. The amount of guidance in the specification is limited because the disclosure identifies possible parameters but does not teach how the event identification circuitry actually obtains or determines several of them. The specification also lacks working examples for determining emotional state, physical condition, and weather as claimed signature parameters. Accordingly, a person of ordinary skill in the art would have to engage in undue experimentation to make and use the full scope of the claimed invention. Accordingly, claims 1–7, 10–13, and 15–18 fail to comply with the enablement requirement of 35 U.S.C. §112(a).
Claim Rejections - 35 USC § 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim 4 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Specifically, Claim 4 recites: "...for each signature in the signature subset, generating... a signature verification likelihood score... wherein the aggregated signature verification likelihood score is generated based on the signature verification likelihood score.".
Because the claim requires generating a score for each signature in the subset, a plurality of scores are generated. However, the end of the claim refers back to a singular "the signature verification likelihood score". It is unclear if the aggregated score is based on a single, specific score from the subset (and if so, which one), or if it is based on all of "the signature verification likelihood scores" collectively. This lack of clear antecedent basis and grammatical agreement renders the scope of the claim indefinite.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1–7, 10–13, and 15–18 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to a judicial exception without significantly more.
This rejection has been made in accordance with the current USPTO subject matter eligibility framework, including MPEP §§ 2103–2106.07, the 2019 Revised Patent Subject Matter Eligibility Guidance, the October 2019 Patent Eligibility Guidance Update, the 2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence, the July 2024 AI Subject Matter Eligibility Examples, the August 4, 2025 USPTO memorandum titled “Reminders on evaluating subject matter eligibility of claims under 35 U.S.C. § 101,” and the USPTO’s guidance concerning Ex parte Desjardins, Appeal No. 2024-000567.
The claims have been evaluated under the broadest reasonable interpretation, and the claims have been considered as a whole.
Step 1: Statutory category
Independent claim 1 is directed to a method and therefore falls within the statutory category of a process. Independent claim 10 is directed to an apparatus comprising event identification circuitry, communications hardware, and signature management circuitry and therefore falls within the statutory category of a machine. Independent claim 15 is directed to a computer program product comprising a non-transitory computer-readable storage medium and therefore falls within the statutory category of a manufacture. Accordingly, the analysis proceeds to Step 2A.
Step 2A, Prong One (Judicial exception)
Independent claim 1 recites: detecting a user authentication event; receiving a signature from the user during the event; determining whether the event corresponds to an authentic event type; determining a signature parameter set for the signature; and, in an instance in which the event corresponds to an authentic event type, storing the signature and corresponding signature parameter set in a signature profile associated with the user within a signature corpus. These limitations recite an abstract idea, namely collecting authentication/ signature information, evaluating whether an event is authentic, determining descriptive signature information, and organizing/ storing the signature information in a user profile.
The claim recites information collection, information evaluation, information classification, and data storage. The “signature”, “user authentication event”, “authentic event type”, “signature parameter set”, “signature profile” and “signature corpus” are used as information items in a data-management process. The claim does not recite an improvement to the way signatures are captured, digitized, encoded, segmented, stored, compressed, transmitted, or cryptographically protected. Rather, the signature and event information are collected, evaluated, and stored for later authentication use.
The claim is similar in character to claims that courts have found abstract where the focus is collecting information, analyzing the information, and presenting or acting on the results of the analysis. In Electric Power Group, LLC v. Alstom S.A., the Federal Circuit recognized claims directed to collecting and analyzing information as abstract. The present claims similarly collect signature/ authentication information, analyze whether an authentication event corresponds to an authentic event type, determine associated signature information, and store the result.
Claims 2–5, 11–13, and 16–18 further recite receiving a signature verification request comprising a candidate signature; determining a signature verification result using a signature authentication model; generating a signature subset; generating signature similarity scores; generating signature verification likelihood scores; generating an aggregated signature verification likelihood score; comparing the aggregated score to a threshold; and determining whether the candidate signature is verified or non-verified. These limitations recite additional abstract data-analysis steps, including comparing information, generating scores, aggregating scores, applying a threshold, and classifying a candidate signature based on the result. The score-generating and threshold-comparison limitations additionally recite mathematical concepts and mental processes. Signature comparison has historically been performed by humans by visually comparing a candidate signature to known signatures. The claims merely recite performing such comparison and decision-making using a signature authentication model, similarity scores, likelihood scores, an aggregated score, and a threshold, without reciting a particular technological improvement to the model or to the computer performing the comparison.
Claims 6–7 further recite identifying an action criteria set comprising one or more requirements for an action request type, determining whether the requirements are satisfied based on the candidate signature, causing performance of one or more operations when the requirements are satisfied, and causing supplemental operations when the requirements are not satisfied. These limitations recite rules-based authorization and decision-making based on the result of the signature analysis.
The claim is also consistent with the reasoning of AI Visualize, Inc. v. Nuance Communications, Inc., where the Federal Circuit looked to the character of the claims as a whole and affirmed ineligibility where the claims were directed to obtaining, manipulating, and using information at a high level of generality rather than to a specific improvement in computer functionality. Here, the character of the elected claims as a whole is signature/ authentication data collection, comparison, scoring, thresholding, and authorization logic, not an improvement to signature-capture technology, biometric-processing technology, computer memory, network operation, cryptographic authentication, or machine-learning technology itself. Independent claim 10 recites substantially the same abstract idea in apparatus form using event identification circuitry, communications hardware, and signature management circuitry. Independent claim 15 recites substantially the same abstract idea in computer-program-product form using instructions stored on a non-transitory computer-readable storage medium. Merely implementing the same abstract signature-management and verification process using generic computer components does not avoid the judicial exception.
Accordingly, claims 1–7, 10–13, and 15–18 recite an abstract idea under Step 2A, Prong One.
Step 2A, Prong Two (Practical Application)
The additional elements, considered individually and in combination, do not integrate the abstract idea into a practical application.
The recited “event identification circuitry”, “communications hardware”, “signature management circuitry”, “signature”, “user authentication event”, “authentic event type”, “signature parameter set”, “signature profile”, “signature corpus”, “signature verification request”, “candidate signature”, “signature authentication model”, “signature subset”, “signature similarity score”, “signature verification likelihood score”, “aggregated signature verification likelihood score”, “threshold”, “action criteria set”, “action request type”, “processor”, “memory”, and “non-transitory computer-readable storage medium” amount to data objects, generic computer components, and generic computer implementation of the abstract signature-management and signature-verification concept.
The claims do not recite a particular improvement to signature-capture technology. They do not improve how a signature is sensed, sampled, imaged, digitized, segmented, filtered, encoded, compressed, encrypted, or transmitted. The claims merely require receiving a signature captured during a user authentication event. The claims also do not recite a particular improvement to biometric-authentication technology. They do not recite a new biometric feature extractor, a new handwriting-recognition architecture, a new sensor arrangement, a new pressure or movement sampling technique, or a new technical process for extracting signature features. Rather, the claims use signature information as input data for comparison, scoring, and verification.
The claims further do not recite a particular improvement to artificial-intelligence or machine-learning technology. The claims do not train a model, update model parameters, modify model architecture, reduce model complexity, reduce model storage, preserve prior model knowledge, improve inference speed, improve model accuracy by a specific claimed technical mechanism, or otherwise improve how the claimed “signature authentication model” operates. The “signature authentication model” is recited functionally as a tool for determining verification results, generating scores, and selecting or comparing signatures. This analysis is consistent with the USPTO’s 2024 AI subject matter eligibility guidance and AI examples, which emphasize that AI-related claims may be eligible when they recite a specific technological improvement or otherwise integrate a judicial exception into a practical application. The present claims do not recite such a specific technological improvement. Instead, the claims use generic computer operations to collect signature information, store signature information, compare a candidate signature to stored signature information, generate scores, compare a score to a threshold, and output or act on the verification result.
This case is distinguishable from Ex parte Desjardins. In Desjardins, the claims were found to reflect an improvement in machine-learning technology itself, including training a machine-learning model on a series of tasks while preserving prior knowledge and reducing complexity/storage burdens. Here, the claims do not recite a particular training technique, model architecture, parameter-update mechanism, memory-saving arrangement, or data structure that improves operation of a machine-learning model. The claimed signature profile, signature subset, similarity score, likelihood score, aggregated likelihood score, and threshold merely determine whether a candidate signature should be verified.
Nor does limiting the abstract idea to the environment of automated signature authentication make the claims eligible. In Recentive Analytics, Inc. v. Fox Corp., the Federal Circuit rejected the argument that applying machine learning to a new field of use was sufficient for eligibility where the claims did not recite a technical improvement to the machine-learning process itself. Similarly here, applying a signature authentication model to signature verification and continuous signature management is a field-of-use limitation, not an integration of the abstract idea into a practical application.
The providing of a signature verification response is also insufficient to integrate the exception into a practical application. Providing the response merely outputs the result of the abstract comparison and scoring process. Likewise, claim 6’s recitation of causing performance of one or more operations associated with an action request type merely uses the result of the abstract verification analysis to authorize or perform an action. Claim 7’s recitation of causing supplemental operations when requirements are not satisfied merely recites remedial or follow-up activity after the abstract verification determination.
Accordingly, the claims do not integrate the judicial exception into a practical application under Step 2A, Prong Two.
Step 2B: (Inventive Concept)
The additional elements, considered both individually and as an ordered combination, do not amount to significantly more than the abstract idea. The claims use generic computer components to perform ordinary computer functions, including detecting events, receiving data, determining parameters, storing data, generating scores, comparing scores to thresholds, and transmitting results. These are conventional data-processing and data-routing operations performed using generic computer technology.
The ordered combination also does not provide an inventive concept. The ordered combination follows the abstract idea itself: detect an event, receive a signature and parameters, determine if the event is authentic, store the signature, receive a candidate signature, select a subset of stored signatures, generate an aggregated likelihood score using a model, compare the score to a threshold, and provide a result based on the comparison. This is no more than the abstract idea implemented on generic computer components.
Dependent claims 2–7 (and counterparts 11–13, 16–18) recite additional steps of requesting verification, using a model to generate subset scores, aggregated scores, and applying thresholds, as well as enforcing action criteria. These limitations merely specify the mathematical weighting, scoring mechanics, and generic business logic used in the abstract evaluation and do not add significantly more.
Claims 10–13 and 15–18 recite apparatus and computer program product counterparts using circuitry, processors, and memories to perform substantially the same operations as method claims 1–7. The recitation of generic circuitry, hardware, and memories does not transform the abstract idea into patent-eligible subject matter.
Accordingly, claims 1–7, 10–13, and 15–18 are directed to a judicial exception without significantly more and are therefore rejected under 35 U.S.C. § 101.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1–2, 6–7, 10–11 and 15–16 are rejected under 35 U.S.C. §102(a)(1) as being anticipated by Hu (Hu et al, US 2007/0177773 A1, 2007).
Regarding claim 1, Hu teaches a method for providing continuous signature management, the method comprising:
detecting, by an event identification circuitry, an occurrence of a user authentication event associated with a user;
( [0017], [0021], [0028]: Hu teaches “rolling enrollment,” where enrollment is performed while the user is normally signing in to perform a transaction, and where the system collects samples during normal signature verification sessions, "The signature… along with an identifier (ID) of the user… are used to verify the user's identity, such that some transaction can be performed or completed". Hu further teaches that the regular authentication process includes submitted samples used for authenticating the user. )
receiving, by communications hardware, a signature from the user captured during the user authentication event;
( [0020], [0021], [0024], [0030], [0032]: Hu teaches a signature verification system including a digitizing tablet, input/output devices, processor, memory, and network interface. Hu further teaches that a user signs his or her name on the digitizing tablet using a stylus, that the tablet records stylus movements during writing, and that, during the verification process, the subject submits a signature sample. )
determining, by the event identification circuitry, whether the user authentication event corresponds to an authentic event type; and
( [0029], [0035], [0036], [0037]: Hu teaches that each submitted sample is compared to reference signatures or a model/ template derived from the reference signatures, and a score is produced. If the score is higher than a pre-selected threshold, authentication succeeds. Hu further teaches that, where sign-on attempts fail but the user passes an ID check, the signer is deemed genuine, and that only samples provided during successful authentication sessions are counted as genuine. )
in an instance in which the user authentication event corresponds to an authentic event type:
determining, by the event identification circuitry, a signature parameter set for the signature, and
( [0029], [0032], [0036], [0042]: Hu teaches that, during the regular authentication process, each submitted signature sample is compared to reference signatures, or to a model or template derived from those reference signatures, and a score is produced indicating how well the new sample matches the reference samples stored in the system. Hu further teaches that the subject submits a signature sample, that the sample is compared to existing reference samples, and that a sample score is computed using a known signature comparison and score-generation process. Hu also teaches that only samples provided during successful authentication sessions are considered genuine, and that the signature information may be evaluated in score space or feature space, where a feature is a multidimensional representation of the characteristics of a signature [corresponding to determining signature-characteristic information, such as a score and/or feature information, for the signature in an instance in which the submitted signature corresponds to an authentic/genuine authentication event]. )
storing, by signature management circuitry, the signature and corresponding signature parameter set in a signature profile associated with the user within a signature corpus.
( [0011], [0017], [0029], [0036-0037], [0039]: Hu teaches updating a reference set of signature samples for the user through selection of one or more obtained signature samples, such that the updated reference set is usable for verifying subsequent signature samples attributed to the user. Hu further teaches dynamically supplementing and updating the initial reference set using samples collected during normal signature verification sessions, adding candidate-reference-pool samples to the reference set when the signer is genuine/ accepted, and saving authentication-related attributes associated with the subject. )
Regarding claim 2, Hu teaches the method of claim 1, further comprising:
receiving, by the communications hardware, a signature verification request, wherein the signature verification request comprises a candidate signature from the user;
( [0021], [0029], [0032]: Hu teaches that a user signs his or her name on a digitizing tablet using a stylus, that the signature and user identifier are used to verify the user’s identity, and that samples are collected during normal signature verification sessions. Hu further teaches that, in the regular authentication process, each submitted sample is compared to reference signatures for the user, and that the subject submits a signature sample in step 206. )
determining, by the signature management circuitry and using a signature authentication model, a signature verification result for the candidate signature based on the candidate signature and the signature profile, wherein the signature verification result is indicative of whether the candidate signature is verified or non-verified; and
( [0028-0029], [0032], [0035]: Hu teaches that rolling enrollment updates an initial reference set used to verify a user. Hu further teaches that each submitted sample is compared to reference signatures, or to a model or template derived from those reference signatures, and that a score is produced indicating how well the new sample matches the reference samples stored in the system. If the score is higher than a pre-selected threshold, authentication succeeds; otherwise, the user is invited to submit another sample and may ultimately be rejected if authentication/ ID-check requirements are not satisfied. )
providing, by the communications hardware, a signature verification response comprising the signature verification result.
( [0021-0024], [0029], [0035]: Hu teaches that input/ output devices may provide the user with information or feedback regarding verification or the transaction, and that signature verification results may be provided to an application server via a network interface. Hu further teaches that the authentication process determines whether authentication succeeds or whether the user is rejected. )
Regarding claim 6, Hu teaches the method of claim 2, wherein the signature verification request further comprises an action request type, wherein the method further comprises:
identifying, by the signature management circuitry, an action criteria set comprising one or more requirements for the action request type;
( [0025-0029], [0040]: Hu teaches that a user’s signature, typically along with a user identifier, is used to verify the user’s identity so that a transaction can be performed or completed, and that the particular transaction depends on the particular application, such as a retail transaction or a banking transaction. Hu teaches that input/ output devices, processor, and memory perform signature verification and possibly part or all of the application-specific transaction, and that signature verification results may be provided via a network interface to an application server that performs the transaction. Hu also teaches an authentication rule requiring the submitted sample to be compared to reference signatures or a model/ template and requiring the resulting score to exceed a pre-selected threshold Tr for authentication to succeed. Hu further teaches application-dependent threshold policy, such as different threshold considerations for retail and banking applications. Thus, Hu teaches identifying requirements for the transaction/ application type, including signature-verification score and threshold requirements, which correspond to an action criteria set for the action request type. )
determining, by the signature management circuitry and based on the candidate signature, whether the one or more requirements are satisfied; and
( [0031-0033]: during the regular authentication process, each submitted signature sample is compared to reference signatures, or to a model or template derived from those reference signatures, and a score is produced indicating how well the new sample matches the stored reference samples. Hu further teaches that if the score is higher than the pre-selected threshold Tr, authentication succeeds; otherwise, the submitted sample is checked against additional threshold criteria, such as lower threshold Tb, or the user may be invited to submit another sample. )
in an instance in which the one or more requirements are satisfied, causing, by the signature management circuitry, performance of one or more operations associated with the action request type.
( [0024], [0028-0031]: Hu teaches that the signature and user identifier are used to verify the user’s identity so that some transaction can be performed or completed, such as a retail transaction or banking transaction. Hu also teaches that when the submitted signature score is higher than the pre-selected threshold Tr, authentication succeeds. Hu further teaches that the processor and memory perform the computations necessary to accomplish signature verification and even the transaction, and that signature verification results may be provided to an application server via the network interface, which performs the transaction. )
Regarding claim 7, Hu teaches the method of claim 6, further comprising in an instance in which one or more of the one or more requirements fail to be satisfied, causing, by the signature management circuitry, performance of one or more supplemental operations associated with the action request type.
( [0031], [0033-0036]: Hu teaches that if the submitted signature sample does not satisfy the acceptance threshold Tr, the user is invited to submit another sample, which is checked against the reference samples again, and this process continues until the maximum number of trials has been exceeded. Hu further teaches that if the submitted sample has a score equal to or lower than Tr, additional processing is performed by comparing the score to a lower threshold Tb and either adding the sample to a candidate reference pool or discarding it as an outlier. If the signer is rejected via ID comparison, no reference-set updating is performed. If the signer failed all sign-on attempts but is accepted after a positive ID check, there are two implications: (i) the signer is indeed genuine; and (ii) the samples collected in the candidate reference pool represent genuine diversions. FIG. 2 expressly teaches the alternative supplemental outcome when the check fails. )
Regarding claims 10–11 and 15–16, the rationale provided in the rejection of claims 1–2 is incorporated herein. In addition, Hu teaches a computer system using computer programs for signature management [0026]. Accordingly, the method for providing continuous signature management of claims 1-2 corresponds to the apparatus of claim 10–11, as well as the computer program product of claims 15–16, and performs the steps disclosed herein. Therefore, the claims are all rejected.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 3–5, 12–13, and 17–18 are rejected under 35 U.S.C. §103 as being unpatentable over Hu (Hu et al, US 2007/0177773 A1, 2007) in view of Chen (Chen et al, US 2021/0124905 A1, 2021).
Regarding claim 3, with deficiencies of Hu noted in square brackets [ ], Hu teaches the method of claim 2, further comprising:
generating, by the signature management circuitry and using the signature authentication model, a signature reference-set [ subset ] comprising one or more signatures from the signature profile;
( [0017], [0028-0029], [0032], [0037]: Hu teaches dynamically supplementing and updating an initial reference set used to verify a user using signature samples collected during normal signature verification sessions. During the regular authentication process, each submitted signature sample is compared to reference signatures, or to a model or template derived from those reference signatures, for the user stored in the system. Then, the submitted sample is compared to existing reference samples and a sample score is computed. Hu further teaches updating the reference set through selection of one or more signature samples from obtained signature samples so the updated reference set is usable for verifying subsequent signature samples attributed to the user. )
generating, by the signature management circuitry and using the signature authentication model, an aggregated signature verification likelihood score for the candidate signature based on the signature reference-set [ subset ]; and
( [0029], [0032], [0042]: Hu teaches that a score is produced indicating how well the new sample matches the reference samples stored in the system. Hu further teaches that the submitted sample is compared to existing reference samples and that the sample score may be computed using any known signature comparison and score-generation process, including distance-metric-based scoring. Hu also teaches that a score is a number used to compare against a threshold to determine whether to accept a signature. Thus, Hu teaches aggregating a verification likelyhood score for the candidate signature based on the user’s signature reference set/model. )
determining, by the signature management circuitry, whether the aggregated signature verification likelihood score for the candidate signature satisfies an aggregated signature verification likelihood score threshold, wherein (a) the signature verification result is indicative that the candidate signature is verified in an instance in which the aggregated signature verification likelihood score for the candidate signature satisfies the aggregated signature verification likelihood score threshold and (b) the signature verification result is indicative that the candidate signature is non-verified in an instance in which the aggregated signature verification likelihood score for the candidate signature fails to satisfy the aggregated signature verification likelihood score threshold.
( [0029], [0032-0033], [0035], [FIG. 2, step 208]: Hu teaches that a score is a single number used to compare against a threshold to determine whether to accept a signature. If the score is higher than a pre-selected threshold Tr, authentication succeeds. Otherwise, the user is invited to submit another sample, and if the maximum number of trials is exceeded, the user is subjected to an ID check and accepted if the ID check is passed or rejected otherwise. )
As noted in square brackets [ ]: Hu’s closest concept is its reference set of signature samples. Hu’s reference set maps better to applicant’s signature profile/ reference collection, not necessarily to applicant’s signature subset as described in the spec. For that reason, Chen is cleaner for the exact subset-selection concept:
generating, by the signature management circuitry and using the signature authentication model, a signature subset comprising one or more signatures from the signature profile;
( [0028-0029], [0036-0037], [0041], [0094]: Chen teaches that a signature verification model receives a target signature and a plurality of genuine signatures associated with the user, where the genuine signatures are stored in storage as user information and may include genuine signatures indexed by user. Chen further teaches that additional signatures may be accepted as genuine signatures and stored in storage, and that the model may be trained or calibrated with all known genuine signatures for the user. Chen also teaches that the signature verification model receives a target signature and a plurality of genuine signatures, such as genuine signature 1 and genuine signature 2, and classifies the target signature based on the plurality of genuine signatures. )
generating, by the signature management circuitry and using the signature authentication model, an aggregated signature verification likelihood score for the candidate signature based on the signature subset; and
( [0003], [0021], [0072-0074], [0094-0096], [0108-0109], [0117-0118]: Chen teaches that a signature verification model pipeline extracts features from a target signature and a genuine signature, encodes and submits both to a neural network to generate a similarity score, and repeats this process for each genuine signature. Chen further teaches that classifier 438 generates a set of similarity scores, where each score is based on similarities between encoded target features and encoded genuine features, and a similarity score is generated for each vector in the plurality of genuine signature vectors relative to the target signature vector. Chen teaches that, given a signature to be verified and N genuine signatures, there may be N similarity scores S1, S2, S3, . . . , SN , and target signature 402 may be classified as genuine when at least one similarity score meets or exceeds threshold T, e.g., max{ S1, S2, S3, . . . , SN } ≥ T. Thus, Chen teaches generating a verification score set for the candidate/ target signature based on the plurality of genuine signatures, and using the score set, such as the maximum score satisfying the threshold, as an aggregated signature verification likelihood score for verifying the candidate signature. )
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Hu’s rolling-enrollment signature verification system to use Chen’s plurality-of-genuine-signatures subset and similarity-score-set verification technique because Hu already teaches verifying a submitted signature by comparing it to the user’s stored reference signatures or to a model/template derived from those reference signatures, and Hu expressly permits use of known signature comparison and score-generation processes. Chen provides one such known process by comparing a target signature to each of a plurality of genuine user signatures, generating a set of similarity scores, and verifying the target signature based on the score set and a threshold. A person of ordinary skill would have been motivated to use Chen’s per-genuine-signature scoring in Hu’s system to improve verification reliability by accounting for natural variation among a user’s genuine signatures. The modification would have been no more than applying a known signature-verification scoring technique to Hu’s known rolling-enrollment signature-verification system for its expected purpose, with a reasonable expectation of success because both references use stored genuine/reference signatures and threshold-based signature verification.
Regarding claim 4, Hu [as modified by Chen] teaches the method of claim 3, further comprising, for each signature in the signature subset, generating, by the signature management circuitry and using the signature authentication model, a signature verification likelihood score for the candidate signature, wherein the aggregated signature verification likelihood score is generated based on the signature verification likelihood score.
( Chen, [0072-0074], [0094-0096], [0109]: Chen teaches determining a similarity between target-signature features and genuine-signature features for each of a plurality of genuine signatures, generating a set of similarity scores comprising a similarity score for each determined similarity, and verifying the target signature based on the set of similarity scores; that FNN 440 generates similarity scores S1, S2, S3, . . . , SN, where each score corresponds to a respective genuine signature relative to the target signature, and that the target signature is classified based on the set of scores. Thus, Chen teaches generating, for each signature in the signature subset/ plurality of genuine signatures, a verification likelihood score for the candidate/ target signature, and generating the final aggregated verification determination based on the score set. )
Regarding claim 5, Hu [as modified by Chen] teaches the method of claim 3, further comprising:
generating, by the signature management circuitry, a signature similarity score for each signature stored in the signature profile, wherein the signature similarity score for a signature is based on an inferred similarity between a corresponding signature parameter set for the signature and a signature parameter set for the candidate signature; and
( Chen, [0047-0051], [0072-0074], [0108-0114]: Chen teaches that a signature verification model receives a target signature and a plurality of genuine signatures associated with the user and classifies the target signature based on similarities with the plurality of genuine signatures. Chen further teaches extracting features from the target signature and genuine signatures, including geometric and temporal information such as coordinate and time information, and generating a similarity score for each genuine signature relative to the target signature. Thus, Chen teaches generating a signature similarity score for each stored genuine signature, where the similarity score is based on inferred similarity between the target/ candidate signature features and the genuine/ reference signature features [corresponding to signature parameter sets]. )
selecting, by the signature management circuitry, one or more signatures for the signature subset based on an associated signature similarity score.
( Chen, [0072-0074], [0109-0111], [0117-0118], [0124-0125]: Chen teaches generating a set of similarity scores S1, S2, S3, . . . , SN for a target signature relative to a plurality of genuine signatures, and verifying/classifying the target signature based on that score set. Chen further teaches that the target signature is classified as genuine when one or more similarity scores satisfy a threshold, such as max{ S1, S2, S3, . . . , SN} ≥ T, and classified as forged when the scores fail to satisfy the threshold. Thus, Chen teaches selecting by using one or more genuine signatures based on their associated similarity scores, such as the genuine signatures whose scores satisfy the threshold. )
Regarding claims 12–13, and 17–18, the rationale provided in the rejection of claims 3, 5 is incorporated herein. In addition, Hu [as modified by Chen] teaches a computer system using computer programs for signature management [Hu, 0026]. Accordingly, the method for providing continuous signature management of claims 3, 5 corresponds to the apparatus of claim 12–13, as well as the computer program product of claims 17–18, and performs the steps disclosed herein. Therefore, the claims are all rejected.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KEN KUDO whose telephone number is (571)272-4498. The examiner can normally be reached M-F 8am - 5pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Vincent Rudolph can be reached at 571-272-8243. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
KEN KUDO
Examiner
Art Unit 2671
/KEN KUDO/Examiner, Art Unit 2671
/VINCENT RUDOLPH/Supervisory Patent Examiner, Art Unit 2671
Prosecution Insights