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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's amendment after final filed on June 3, 2026 (herein “Amendment”) has been entered by way of the Request for Continued Examination filed July 8, 2026.
Response to Arguments
Applicant's arguments and amendments regarding the rejection of claims 1, 11 and 20 and pending claims depending therefrom under 35 U.S.C. 101 for being directed towards an abstract idea without a practical application or significantly more have been fully considered but they are not persuasive.
Applicant first argues on pages 7–9 of the Amendment that the claim limitations do not fall within any of the enumerated groupings constituting an abstract idea. Applicant argues that because the claims do not recite any mathematical relations, formulas or calculations, they do not recite an the mathematical concept type of abstract idea, and in particular cites to last year’s subject matter eligibility Memo1 on page 3 stating that “Examiners should be careful to distinguish claims that recite an exception (which require further eligibility analysis) from claims that merely involve an exception (which are eligible and do not require further eligibility analysis).” It is noted that the Memo never defines the “claims that recite an exception” to mean that the claim limitations “must” set forth or describe mathematical relationships, calculations, formulas or equations. Rather, page 3 of the Memo advises Examiners to “consider for example, the published USPTO examples 39” which merely recites “training the neural network in a first stage using the first training set.” At least, the present application does recite mathematical relationships in that it recites “performing one or more operations to update the machine learning model.” These operations would be understood in view of the Specification, for example in ¶¶ 56–58 as performing the denoiser operations according to specific inputs, and where ¶47 teaches the operations of the denoising model directed towards the mathematical concept of a sum of the probability flow of an ordinary differential equation (ODE) and a time-varying Langevin diffusion stochastic differential equation (SDE).
Applicant further argues that the claimed “performing one or more operations to train the machine learning model based on the one or more augmented samples and the one or more augmentation parameters,” do nothing more than encompass a broad array of different techniques for updating a machine learning model, each of which could involve or rely upon mathematical concepts. However, the Examiner notes that the first step in subject matter eligibility analysis is to determine what the broadest reasonable interpretation is for each claim limitations, which according to MPEP 2111.01 is limited to the plain meaning in view of the specification. As noted above, the Specification sets forth that the training is mathematical calculations, not just “involving” and moreover, the training is entirely a mathematical process. Therefore, the Examiner’s finding that the “performing one or more operations to update the machine learning model …” limitations recites the judicial exception of a mathematical concept is maintained herein. The outstanding rejection has not found the above limitation to be directed towards any other judicial exceptions, and therefore, Applicant’s arguments regarding the above limitations as not being directed towards mental processes or organizing human activities are not relevant at this time.
Next, on pages 10–11, Applicant argues that the claims recite limitations that necessarily integrate any purported abstract idea into a practical application. Here, Applicant again references Ex Parte Desjardins2, but as already responded to in the Final Action issued April 8, 2026 (herein “Final Action”), the facts and claim limitations at issue in Ex Parte Desjardins are distinguished from those of the present application, and therefore, inferences as to what constitutes subject matter eligibility from the Ex Parte Desjardins case are not applicable to the present application. Similarly, the Final Action on pages 3–4 already addressed Applicant’s arguments regarding there being a practical application recited per MPEP 2106.04(d)(1). The present amendments to the claims still do not recite a practical application.
Therefore, in view of the above, while all of Applicant's arguments and amendments regarding the rejection of claims 1, 11 and 20, and pending dependent claims therefrom under 35 U.S.C. 101 as being directed to an abstract idea without a practical application or significantly more have been fully considered, they are not persuasive, and the rejection is substantively maintained, with updates to the rejection made below to reflect the present amendments to the claims.
Applicant's arguments and amendments in the Amendment regarding the rejection of claims 1, 11 and 20 on the ground of non-statutory double patenting have been fully considered and are persuasive. The rejection against claims 1, 11 and 20 on the ground of non-statutory double patenting is withdrawn.
Applicant’s arguments and amendments in the Amendment with respect to the rejection of claims 1, 11 and 20 and various dependent claims therefrom under 35 U.S.C. 102(a)(2) have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made under 35 U.S.C. 103 in view of Lecue et al., US Patent No. 10,565,475.
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, 3–11, and 13–20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without a practical application or significantly more.
Regarding claims 1, 11 and 20, these claims recite the following limitations which are found to be abstract ideas not reciting a practical application or significantly more, with claim 1 being exemplary:
receiving training data that includes one or more samples; generating one or more augmented samples based on the one or more samples and one or more augmentation parameters (abstract idea as a mental process per MPEP §2106.04(a)(2)(III) as a human mind is capable of reading/seeing (receiving) data samples and at least with pen and paper, altering/augmenting a sample).
performing one or more operations to update the machine learning model based on providing the one or more augmented samples along with the one or more augmentation parameters as input to the machine learning model to generate a trained machine learning model (abstract idea as mathematical concepts per MPEP §2106.04(a)(2)(I), as updating a machine learning model is well-known to be a mathematical process, and as well, per the BRI of these limitations which includes the plain meaning in view of the Specification).
Claims 1, 11 and 20 further recite additional elements: claim 20 is directed towards a system comprising one or more memories, and one or more processors, claim 11 is directed towards a non-transitory computer-readable media and claim 1 recites “computer-implemented.” While a non-transitory computer-readable media of claim 11, the computer-implemented limitation of claim 1, and the processor(s) and memor(ies) of claim 20 are additional elements, they are not sufficient to recite a practical application of the abstract ideas recited in claims 1, 11 and 20 as they amount to mere generic computer elements and thus amount to no more than a recitation of the words "apply it" (or an equivalent) or are no more than mere instructions to implement an abstract idea or other exception on a computer. see MPEP §2106.05(f).
Further, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because when considered separately and in combination, the above recited additional elements from claims 1, 11 and 20 do not add significantly more (also known as an “inventive concept”) to the exception. Rather, the additional elements disclosed above perform well-understood, routine, conventional computer functions as recognized by the court decisions listed in MPEP § 2106.05(d).
Therefore, independent claims 1, 11 and 20 are directed towards an abstract idea without a practical application or significantly more.
Regarding claims 3 and 13, the limitations are merely directed towards further abstract ideas, specifically mental concepts per MPEP §2106.04(a)(2)(I), as the human mind can have input information regarding data augmentation, and additional samples.
Regarding claims 4–6, 14 and 15, the limitations are merely directed towards data aspects of the previously recited abstract ideas in the independent claim, but do not add significantly more or constitute a practical application.
Regarding claims 7 and 16, the limitations are merely directed towards further abstract ideas, specifically mental concepts per MPEP §2106.04(a)(2)(I), as the human mind can add corruption to samples, at least using pen and paper.
Regarding 8 and 17, the limitations are merely directed towards characteristics of the previously recited abstract ideas in the independent claim, but do not add significantly more or constitute a practical application.
Regarding claims 9 and 19, the limitations are merely directed towards further abstract ideas, specifically mental concepts per MPEP §2106.04(a)(2)(I), as the human mind can reduce corruption to samples, at least using pen and paper. The additional element of “using the trained machine learning model” does not integrate the abstract idea into a practical application because of the high level of generality recited by the limitation. Nor does the claimed machine learning model recite significantly more.
Regarding claim 10, the limitations are merely directed towards further abstract ideas, specifically mental concepts per MPEP §2106.04(a)(2)(I), as the human mind can determine parameters.
Regarding claim 18, the limitations are merely directed towards characteristics of the previously recited abstract ideas in the independent claim, but do not add significantly more or constitute a practical application.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 4–6, 8, 10–11, 14–15, 17–18 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al., US Patent Application Publication No. US 2023/0196718 A1 (herein “Li”) in view of Lecue et al., US Patent No. 10,565,475 B2 (herein “Lecue”).
Regarding claims 1, 11 and 20, with claim 1 as exemplary, substantive differences between the claims noted in curly brackets {}, and deficiencies of Li noted in square brackets [], Li teaches { A computer-implemented method for training a machine learning model, the method comprising - claim 1 / One or more non-transitory computer-readable media storing instructions that, when executed by at least one processor, cause the at least one processor to perform the steps of - claim 11 / A system, comprising: one or more memories storing instructions; and one or more processors that are coupled to the one or more memories and, when executing the instructions, are configured to - claim 20} (Li Abstract, ¶ 20, an image augmentation device including a memory and processor configured for performing operations, the memory being a hard disk, therefore non-transitory computer readable media):
receiving training data that includes one or more samples (Li ¶¶ 33 and 43, first object contour is extracted from (receiving) a first image mask, where the first object contour is used to generate a sample image for performing machine learning, and therefore the first object contour is training data);
generating one or more augmented samples based on the one or more samples and one or more augmentation parameters (Li ¶35, the first object contour is superimposed to a superimposed region in a second image mask according to the augmentation parameter to generate a third image mask (augmented samples)); and
performing one or more operations to update the machine learning model based on providing the one or more augmented samples [along with the one or more augmentation parameters as input to the machine learning model] to generate a trained machine learning model (Li ¶¶ 31,43, fig. 1, sample image IMG is generated according to the first object contour (one or more augmented samples) for performing machine learning, where fig. 1 shows IMG as an input to machine learning model MLM, and the MLM is trained (update) based on the corresponding image samples from the new image masks).
Li does not explicitly teach where Lecue teaches along with the one or more augmentation parameters as input to the machine learning model (Lecue figs. 1C and 1E, col. 4, ll. 8–12, 39–67 and col. 5, ll. 25–51, augmented data sequences 145 are input to the machine learning model to train the machine learning model, where the augmented data sequences include physical property data (augmentation parameters) including deformation and rotation data for the objects, as well as the objects that are to have the physical property data applied).
Therefore, taking the teachings of Li and Lecue together as a whole, it would have been obvious to a person having ordinary skill in the art (herein “PHOSITA”) before the effective filing date of the claimed invention to have modified the machine learning of Li to include the augmented data sequences as input for training including the object image and the physical property data as disclosed in Lecue at least because doing so would provide a way to train machine learning models using limited training data while still having the training data represent realistic modifications/permutations of objects, thus resulting in more scalable and applicable machine learning models across different domains. See Lecue col. 3, ll. 16–40, and col. 6, ll. 44–57.
Regarding claims 4 and 14, Li teaches wherein each augmentation parameter included in the one or more augmentation parameters indicates at least one of a geometric transformation, a color change, a filtering, a masking, a cropping, a compression, a quantization, a pixelation, a decimation, a composition, a cutout, a cutmix, or a mixup (Li ¶¶ 35 and 40, the augmentation parameter as a scaling parameter, contour moving distance, and a contour rotation angle and a range for superimposition, all of which are geometric transformations).
Regarding claim 5, Li teaches wherein each augmentation parameter included in the one or more augmentation parameters indicates at least one of an isotropic scaling, an anisotropic scaling, a rotation, an integer translation, a fractional translation, or a flip along an axis (Li ¶¶ 35 and 40, the augmentation parameter as a scaling parameter, contour moving distance, and a contour rotation angle (a rotation) and a range for superimposition, all of which are geometric transformations).
Regarding claims 6 and 15, Li teaches wherein the one or more operations to update the machine learning model are further based on the one or more samples (Li fig. 1, ¶ 43, operations that generate the sample image IMG used for performing machine learning, are according to the first object contour and the second object contour (the one or more samples)).
Regarding claims 8 and 17, Li teaches wherein the trained machine learning model comprises a generative model (Li ¶ 27, fig. 1, the machine learning model executed/generated by the processor including a generative adversarial network (GAN)).
Regarding claim 10, Li teaches further comprising determining the one or more augmentation parameters (Li ¶ 42, the augmentation parameter can be adjusted (determining) according to the relationship between various object types and contours of various objects).
Regarding claim 18, Li teaches wherein the trained machine learning model comprises at least one of a generative adversarial network, a U-Net architecture, a transformer architecture, a vision transformer architecture, a recurrent interface network architecture, or a convolutional architecture (Li ¶ 27, fig. 1, the machine learning model executed/generated by the processor including a generative adversarial network (GAN)).
Claims 3 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Li and Lecue, as set forth above in the independent claims, and further in view of Luo et al., US Patent Application Publication No. US 2023/0273914 A1 (herein “Luo”).
Regarding claims 3 and 13, with claim 3 as exemplary, Li teaches comprising performing one or more operations on another sample using the trained machine learning model by inputting, into the trained machine learning model, the another sample (Li fig. 1, ¶¶33–36, mask augmentation model performing mask augmentation upon inputted image masks MSK(1)-MSK(N), thus other first image masks from the set of MSK(1)-MSK(N) being another sample). Li does not explicitly teach, where Luo teaches and an indication of no augmentation (Luo ¶¶ 356–357 and fig. 15, data augmentation performed upon a sample according to an intensity value from a Gaussian distribution, including zero values, thus zero intensity being an indication of no augmentation).
Therefore, taking the teachings of Li and Luo together as a whole, it would have been obvious to a person having ordinary skill in the art (herein “PHOSITA”) before the effective filing date of the claimed invention to have modified the augmentation of Li to include augmentation levels according to a Gaussian distribution including no augmentation as disclosed in Luo at least because doing so would improve sample diversity and thus provide training data that will avoid low precision and a low generalization capability in a model obtained through training with the training data. See Luo ¶¶ 4 and 8.
Claims 7 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Li and Lecue, as set forth above in the independent claims, and further in view of van Walsum et al., US Patent Application Publication No. US 2020/0222018 A1 (herein “Walsum”).
Regarding claims 7 and 16, with claim 7 as exemplary, while Li teaches performing operations to the one or more augmented samples (Li fig. 1, ¶ 43, the third image mask is input to the GAN model and operated on to produce an sample image IMG), Li does not explicitly teach the operations “to add corruption.” Walsum teaches to add corruption (Walsum ¶202, in addition to the geometric transformations, the data augmented can have added Gaussian noise to the pixel values).
Therefore, taking the teachings of Li and Walsum together as a whole, it would have been obvious to a PHOSITA before the effective filing date of the claimed invention to have modified the augmentation of Li to include adding Gaussian noise as corruption as disclosed in Walsum at least because doing so would make the trained model robust to noise. See Walsum ¶ 202.
Claims 9 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Li and Lecue, as set forth above in the independent claims, and further in view of Bhattacharya et al., US Patent No. US 12,444,051 B2 (herein “Bhattacharya”).
Regarding claims 9 and 19, with claim 9 as exemplary, Li does not explicitly teach, but Bhattacharya teaches performing one or more operations to reduce corruption in a content item using the trained machine learning model (Bhattacharya fig. 19, col. 18, l. 62–65, and col. 19, l. 7–10, image is input to the trained neural network, and a denoised image is produced as one or more operations of the trained neural network).
Therefore, taking the teachings of Li and Bhattacharya together as a whole, it would have been obvious to a PHOSITA before the effective filing date of the claimed invention to have modified the machine learning system of Li to include operations that reduce noise/corruption as disclosed in Bhattacharya at least because doing so would provide an improved tool for reducing noise in an image, and thus improve the imaging capability of an existing imaging system. See Bhattacharya col. 6, ll. 46–48 and col. 1, ll. 55–59.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHELLE M KOETH whose telephone number is (571)272-5908. The examiner can normally be reached Monday-Thursday, 09:00-17:00, Friday 09:00-13:00, EDT/EST.
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.
MICHELLE M. KOETH
Primary Examiner
Art Unit 2671
/MICHELLE M KOETH/Primary Examiner, Art Unit 2671
1 “Reminders on evaluating subject matter eligibility of claims under 35 U.S.C. 101,” Deputy Commissioner Charles Kim, USPTO (August 4, 2025) (available at: https://www.uspto.gov/sites/default/files/documents/memo-101-20250804.pdf) (herein “the Memo”).
2 Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision).
Prosecution Insights