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 .
Status of Claims
This Office Action is in response to the communication filed on 30 May 2024.
Claims 1-20 are being considered on the merits.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 30 August 2024, 22 August 2025, 16 January 2026, and 13 July 2026 have been considered. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, initialed and dated copies of Applicant's IDS forms 1499 are attached to the instant Office action.
Claim Rejections - 35 USC § 112
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.
Claims 1, 3, 8, and 10 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.
Claims 1 and 8 recite the limitation "the input image of the second user" in the third to last claim limitation of the respective claims.
Claims 3 and 10 recite the limitation "the input image of the second user" in the second claim limitation of the respective claims.
There is insufficient antecedent basis for this limitation in each of the claims.
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 1-6, 8-13, and 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Mequanint, et al (US 2020/0082062 A1; hereinafter “Mequanint”) in view of Goldwerger, et al (US 2026/0161762 A1; hereinafter “Goldwerger”)
Claims 1, 8 and 15:
A system, comprising: at least one processor programmed or configured to (Mequanint, para. 0008: “In another example, a non-transitory computer-readable medium is provided that has stored thereon instructions that, when executed by one or more processors, cause the one or more processor to”): generate a plurality of image feature templates using a first machine learning model, (Goldwerger, para. 0042: “ The sample may be analyzed and converted into a biometric template using one or more template generators (e.g., machine learning model) that transform the biometric samples into a biometric template.”)
A method, comprising: generating, with at least one processor (Mequanint, para. 0008: “In another example, a non-transitory computer-readable medium is provided that has stored thereon instructions that, when executed by one or more processors, cause the one or more processor to”): a plurality of image feature templates using a first machine learning model, (Goldwerger, para. 0042: “ The sample may be analyzed and converted into a biometric template using one or more template generators (e.g., machine learning model) that transform the biometric samples into a biometric template.”)
A computer program product, the computer program product comprising at least one non-transitory computer-readable medium including one or more instructions that, when executed by at least one processor, cause the at least one processor to (Mequanint, para. 0008: “In another example, a non-transitory computer-readable medium is provided that has stored thereon instructions that, when executed by one or more processors, cause the one or more processor to”): generate a plurality of image feature templates using a first machine learning model, (Mequanint, para. 0039: “Face authentication, for example, can compare a face of a device user in an input image with known features (e.g., stored in one or more templates) of the person the user claims to be, in order to authenticate that the user of the device is, in fact, the person. A similar process can be performed for fingerprint authentication, voice authentication, and other biometric-based authentication methods.”)
wherein the first machine learning model is configured to authenticate an identity of one or more first users based on an input image of the one or more first users, (Mequanint, para. 0002: “ Face authentication, for example, can compare a face of a device user in an input image with known features of the person the user claims to be, in order to authenticate that the user of the device is, in fact, the person”)
wherein each image feature template of the plurality of image feature templates is associated with a positive authentication of the identity of the one or more first users (Mequanint, para. 0039: “Face authentication, for example, can compare a face of a device user in an input image with known features (e.g., stored in one or more templates) of the person the user claims to be, in order to authenticate that the user of the device is, in fact, the person. A similar process can be performed for fingerprint authentication, voice authentication, and other biometric-based authentication methods.”) during a time interval, (Mequanint, para. 0044: “Another possible solution is to perform multiple re-enrollments with some fixed interval of time.”)
wherein each image feature template is a multi-dimensional vector (Goldwerger, para. 0024: “In some embodiments, a biometric measurement and biometric template can each be translated into a vector in multi-dimensional Euclidian space.”), where dimensions of the multi-dimensional vector include values that are representative of features of an image, and (Goldwerger, para. 0042: “A biometric template may refer to a digital representation of unique features or characteristics of a person (e.g., user) that have been extracted from one or more biometric samples”)
wherein, when generating the plurality of image feature templates, the at least one processor is programmed or configured to: generate each image feature template of the plurality of image feature templates (Mequanint, para. 0055: “For example, during enrollment (which can also be referred to as registration), an owner of the computing device and/or other user with access to the computing device can input one or more biometric data samples (e.g., an image, a fingerprint sample, a voice sample, or other biometric data), and representative features of the biometric data can be extracted by the feature extraction engine 204. The representative features of the biometric data can be stored as one or more templates in the template storage 208. Using images as an example of biometric data, several images can be captured of the owner or user with different poses, positions, facial expressions, lighting conditions, and/or other characteristics.”) for the one or more first users for a point in time of the time interval (Mequanint, para. 0044: “Another possible solution is to perform multiple re-enrollments with some fixed interval of time.”) based on one or more input images of the one or more first users (Mequanint, para. 0038: “As noted above, object authentication or verification systems can be used to authenticate or verify objects. For example, using face authentication as an example, a input query face image can be compared with stored or enrolled representations of a person's face. In general, face authentication needs higher recognition accuracy since it is often related to access control of a device or system”) received during the time interval that resulted in a positive authentication of the identity of the one or more first users during the time interval; (Mequanint, para. 0044 and 0045: “Another possible solution is to perform multiple re-enrollments with some fixed interval of time.” “One alternative is to add to the existing templates only when the similarity score of a presented biometric representation exceeds an authentication threshold, which can be a high threshold to ensure the probability of an impostor being authenticated is low”)
generate a second machine learning model based on the plurality of image feature templates; (Goldwerger, para. 0093: “At stage 304, the biometric measurements are used as input to respective machine learning models. The machine learning model can translate the input from one domain (e.g., images) to another domain (e.g., vectors).”)
generate a predicted image feature template using the second machine learning model, (Mequanint, para. 0037: “Using face identification as an illustrative example of object identification, an enrolled database containing the features of enrolled faces can be used for comparison with the features of one or more given query face images (e.g., from input images or frames). The enrolled faces can include faces registered with the system and stored in the enrolled database, which contains known faces.”) wherein the predicted image feature template comprises an image feature template that is based on a predicted image of a second user (Goldwerger, para. 0039: “In some embodiments, the output of the trained machine learning model 165 may include confidence data that indicates a level of confidence that the output (e.g., prediction) is appropriate or true, for instance”) with regard to a future time after the time interval; (Mequanint, para. 0044: “Another possible solution is to perform multiple re-enrollments with some fixed interval of time.” Examiner notes Mequanint teaches users where enrollments occur at fixed intervals of time and where a first user and a second user enrollment denotes a sequence where a first user enrolls at a first time interval and a second user enrolls at a time interval after the first, sequentially, and therefore in the future).
generate a current image feature template based on the input image of the second user received after the time interval; (Mequanint, para. 0044: “The adaptive biometrics approaches attempt to solve the problems by updating the enrolled templates, which can help to incorporate the variations during the test phase into the reference templates…Another possible solution is to perform multiple re-enrollments with some fixed interval of time.” Examiner notes Mequanint teaches generating updated templates with re-enrollments i.e. current as to the re-enrollment time).
determine that the current image feature template corresponds to the predicted image feature template; and perform an action based on determining that the current image feature template corresponds to the predicted image feature template. (Mequanint, para. 0045: “One alternative is to add to the existing templates only when the similarity score of a presented biometric representation exceeds an authentication threshold, which can be a high threshold to ensure the probability of an impostor being authenticated is low”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Goldwerger into Mequanint. Mequanint teaches techniques and systems for authenticating a user of a device; Goldwerger teaches a decentralized biometric identification and authentication network. One of ordinary skill would have been motivated to combine the teachings of Goldwerger into Mequanint in order to preserve the security of biometric templates (Goldwerger, para. 0019).
Claims 2, 9 and 16:
wherein, when generating the plurality of image feature templates using the first machine learning model, the at least one processor is programmed or configured to: extract a first image feature template from one or more layers of the first machine learning model after training the first machine learning model. (Mequanint, para. 0056: “For example, a neural network can be applied to an input image including a face of a person, and can learn the distinctive features of the face. The neural network can be a classification network including hidden convolutional layers that apply kernels (also referred to as filters) to the input image to extract features.”)
Claims 3, 10, and 17:
add the input image of the second user to the plurality of images of the one or more first users (Mequanint, para. 0037: “Using face identification as an illustrative example of object identification, an enrolled database containing the features of enrolled faces can be used for comparison with the features of one or more given query face images (e.g., from input images or frames). The enrolled faces can include faces registered with the system and stored in the enrolled database, which contains known faces. An enrolled face that is the most similar to a query face image can be determined to be a match with the query face image.” Examiner notes Mequanint teaches a multi-user device where a dataset of faces includes multiple users including a first and second user) in a training dataset to provide an updated training dataset; and (Mequanint, para. 0112: “In an example in which the deep learning network 800 is used to identify features in images, the network 800 can be trained using training data that includes both images and labels for different features. For instance, training images can be input into the network, with each training image having a label indicating the classes of the one or more features in each image (basically, indicating to the network what the features are and what characteristics they have). “)
retrain the first machine learning model based on the updated training dataset. (Mequanint, para. 0111: “The interconnection can have a tunable numeric weight that can be tuned (e.g., based on a training dataset), allowing the deep learning network 800 to be adaptive to inputs and able to learn as more and more data is processed.”)
Claim 4:
The system of claim 3, wherein, when generating the plurality of image feature templates using the first machine learning model, the at least one processor is programmed or configured to: extract a first image feature template (Goldwerger, para. 0018: “ A biometric template may refer to a digital representation of a unique feature or characteristic of a person that has been extracted from one or more biometric samples (e.g., images of a person's face).”) from one or more layers of the first machine learning model after retraining the first machine learning model based on the updated training dataset. (Mequanint, para. 0056: “One illustrative example of a feature extraction process performed by the feature extraction engine 204 that can generate deep learning features is neural network (e.g., a deep learning network) based feature extraction. For example, a neural network can be applied to an input image including a face of a person, and can learn the distinctive features of the face. The neural network can be a classification network including hidden convolutional layers that apply kernels (also referred to as filters) to the input image to extract features.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Goldwerger into Mequanint, as modified, as set forth above with respect to claims 1, 8, and 15.
Claims 11 and 18:
extract a first image feature template from the first machine learning model (Goldwerger, para. 0018: “ A biometric template may refer to a digital representation of a unique feature or characteristic of a person that has been extracted from one or more biometric samples (e.g., images of a person's face).”) after retraining the first machine learning model based on the updated training dataset. (Mequanint, para. 0111: “The interconnection can have a tunable numeric weight that can be tuned (e.g., based on a training dataset), allowing the deep learning network 800 to be adaptive to inputs and able to learn as more and more data is processed.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Goldwerger into Mequanint, as modified, as set forth above with respect to claims 1, 8, and 15.
Claims 5 and 12:
wherein the at least one processor is further programmed or configured to: train the first machine learning model based on a training dataset (Mequanint, para. 0111: “The interconnection can have a tunable numeric weight that can be tuned (e.g., based on a training dataset), allowing the deep learning network 800 to be adaptive to inputs and able to learn as more and more data is processed.”) of a plurality of facial images of the one or more first users. (Mequanint, para. 0004: “Using a face as an example, large intra-class face variations can be due to the pose of the person, removable accessories and/or features (e.g., beards, mustaches, glasses, scarves, or other items), facial expressions, occlusions, aging, change in facial features, lighting conditions, among others.”)
Claims 6 and 13:
wherein the first machine learning model is a convolutional neural network model. (Mequanint, para. 0118 and fig. 9: “One example includes a convolutional neural network (CNN), which includes an input layer and an output layer, with multiple hidden layers between the input and out layers. The hidden layers of a CNN include a series of convolutional, nonlinear, pooling (for downsampling), and fully connected layers. The deep learning network 800 can include any other deep network other than a CNN, such as an autoencoder, a deep belief nets (DBNs), a Recurrent Neural Networks (RNNs), among others.”)
Claims 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Mequanint, et al (US 2020/0082062 A1; hereinafter “Mequanint”) in view of Goldwerger, et al (US 2026/0161762 A1; hereinafter “Goldwerger”) and further in view of Streit (US 2020/0228336 A1; hereinafter “Streit”)
Claims 7 and 14:
The system of claim 1, wherein the second machine learning model is a long short-term memory recurrent neural network model. (Streit, para. 0164: “According to one example, the first neural network (i.e., the generation neural network) can be architected as a Long Short-Term Memory (LSTM) model which is a type of Recurrent Neural Network (RNN).”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Streit into Mequanint as modified. Streit teaches an authentication system with privacy-enabled biometric processing. One of ordinary skill would have been motivated to combine the teachings of Streit into Mequanint as modified in order to enable significant execution efficiency benefits over conventional approaches via incremental re-retraining rather than full retraining (Streit, para. 0137).
Search Notes
PE2E Search best results CPC + key terms: “CNN” “LSTM” “biometric” “authentication” “image” “users” “dataset”
IP.com best search results based on application as main search idea with priority date filter
Google scholar search best results from: CNN LSTM biometric authentication images facial users time interval dataset re-train second [before:2020]
Prior Art
Balogh, et. Al. (US 2021/0152550 A1) teaches using a machine learning system to analyze authentication information.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Sally T. Ley whose telephone number is (571)272-3406. The examiner can normally be reached Monday - Thursday, 10:00am - 6:00pm ET.
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, Viker Lamardo can be reached at (571) 270-5871. 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.
/STL/Examiner, Art Unit 2147
/NHAT HUY T NGUYEN/Primary Examiner, Art Unit 2147
Prosecution Insights