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 .
Drawings
The drawings are objected to because "Unlabeled data clusters" component shown in figures 2A, 3A, 4A, and 5A does not have a designated reference number while having repeated appearances of said component in multiple embodiments. Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. The figure or figure number of an amended drawing should not be labeled as “amended.” If a drawing figure is to be canceled, the appropriate figure must be removed from the replacement sheet, and where necessary, the remaining figures must be renumbered and appropriate changes made to the brief description of the several views of the drawings for consistency. Additional replacement sheets may be necessary to show the renumbering of the remaining figures. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Specification
The disclosure is objected to because of the following informalities:
Paragraph 2, Line 6 “…maximize the monetization online platforms…”
Correction: “…maximize the monetization of online platforms…”
Multiple instances of “weakly labeled training data” are referred to as reference number 260 in the specification, while the drawings refer to “weakly labeled training data” as reference number 270.
Paragraph 35, line 4
Paragraph 36, line 2, line 3, line 5, line 16
Paragraph 38, line 2 “…weakly labeled training data 270…”
Correction: “…weakly labeled training data 260…”
Paragraph 47, line 7 “self-training data generator” is referred to with reference number 450. The drawings refer to “self-training data generator” with reference number 330.
Paragraph 51, line 3 “generative augmentation generator” is referred to with reference number 3430. The appropriate reference number disclosed earlier in the specification is 340.
Appropriate correction is required.
Claim Objections
Claims 1-20 objected to because of the following informalities:
Claim 1: starting on line 5, ending on line 9 is a run-on sentence and is difficult to interpret. The examiner recommends that the applicant corrects the claim such that it reads clearer and more concisely. The applicant may break up the limitation into multiple sentences.
Claim 2: Line 1 “…wherein the generating the weakly labeled…”
Correction: “…wherein generating the weakly labeled…”
Claim 4: Line 1 “…wherein the training the generative augmentation…”
Correction: “…wherein training the generative augmentation…”
Claim 6: Line 1 “…wherein the generating the one…”
Correction: “…wherein generating the one …”
Line 4: “proving the data sample to the…”
Correction: “providing the data sample to the…”
Line 5: “…model, next new data sample…”
Correction: “…model, a next new data sample…”
Claim 8: starting on line 6, ending on line 10 is a run-on sentence and is difficult to interpret. The examiner recommends that the applicant corrects the claim such that it reads clearer and more concisely. The applicant may break up the limitation into multiple sentences.
Claim 9: Line 1 “…wherein the generating the weakly labeled…”
Correction: “…wherein generating the weakly labeled…”
Claim 11: Line 1 “…wherein the training the generative augmentation…”
Correction: “…wherein training the generative augmentation…”
Claim 13: Line 1 “…wherein the generating the one…”
Correction: “…wherein generating the one …”
Line 4: “proving the data sample to the…”
Correction: “providing the data sample to the…”
Line 5: “…model, next new data sample…”
Correction: “…model, a next new data sample…”
Claim 15: starting on line 7, ending on line 12 is a run-on sentence and is difficult to interpret. The examiner recommends that the applicant corrects the claim such that it reads clearer and more concisely. The applicant may break up the limitation into multiple sentences.
Claim 16: Line 1 “…wherein the generating the weakly labeled…”
Correction: “…wherein generating the weakly labeled…”
Claim 17: Line 1 “…wherein the training the generative augmentation…”
Correction: “…wherein training the generative augmentation…”
Claim 19: Line 1 “…wherein the generating the one…”
Correction: “…wherein generating the one …”
Line 4: “proving the data sample to the…”
Correction: “providing the data sample to the…”
Line 5: “…model, next new data sample…”
Correction: “…model, a next new data sample…”
Appropriate correction is required.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 1, 7-8, 14-15, and 20 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 1, 7-8, 14-15, and 20 of copending Application No. 18/487,487 in view of Li et al. ("A Novel Generative Model With Bounded-GAN for Reliability Classification of Gear Safety", hereinafter Li). The copending application (18/487,487) discloses similar limitations shown in the table below to the instant application (18/487,460). The only difference between both applications appear in the second limitation stated in claims 1, 8, and 15 of the instant application (18/487,460) stating “includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels,”. These are underlined in the table below to show the differencs. The second limitation is taught by Li ([Page 8773 Left Co; Page 8772 Abstract; Page 8777 Right Col Para 4] Li discloses generating gear data and using a MCL scheme to label the generated data in view of gear data. Page 8772 discloses in the abstract that the gear data is generated based on existing gear data. Page 8777, right column, paragraph 4, starting with “For each class” discloses integrating the synthetic data with labels into the real data which is used to train classifiers. The integration of the synthetic data into the real data is augmentation. Additionally, on the same page, right column, Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. Parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data. The algorithm ends by integrating the instances into both the real data and synthetic data to add noise, effectively creating augmented data.) It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the similar limitations taught in the corresponding application (18/487,487) with the method for training classifiers for gear data taught by Li. One would be motivated to combine these to improve the training data for classifying and categorization models with multiple methods of augmentation.
This is a provisional nonstatutory double patenting rejection.
Instant Application: 18/487,460
Corresponding Application: 18/487,487
Claim 1: A method, comprising: receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels, a data sample in the supervised training data with a label and a new data sample from the weakly labeled training data with the same label have varying characteristics; obtaining augmented training data based on the supervised training data and the weakly labeled training data; and training, via machine learning, a robust content categorization model based on the augmented training data.
Claim 1: A method, comprising: receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes data samples each of which is assigned with one of the plurality of labels via consistent self-training so that a data sample in the supervised training data with a label and a data sample from the weakly labeled training data with the same label have varying characteristics; obtaining augmented training data based on the supervised training data and the weakly labeled training data; and training, via machine learning, a robust content categorization model based on the augmented training data.
Claim 7: The method of claim 1, further comprising: receiving content to be categorized; and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data.
Claim 7: The method of claim 1, further comprising: receiving content to be categorized; and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data.
Claim 8: A machine-readable medium having information recorded thereon, wherein the information, when read by machine, causes the machine to perform the following steps: receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels, a data sample in the supervised training data with a label and a new data sample from the weakly labeled training data with the same label have varying characteristics; obtaining augmented training data based on the supervised training data and the weakly labeled training data; and training, via machine learning, a robust content categorization model based on the augmented training data.
Claim 8: A machine-readable and non-transitory medium having information recorded thereon, wherein the information, when read by the machine, causes the machine to perform the following steps: receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes data samples each of which is assigned with one of the plurality of labels via consistent self-training so that a data sample in the supervised training data with a label and a data sample from the weakly labeled training data with the same label have varying characteristics; obtaining augmented training data based on the supervised training data and the weakly labeled training data; and training, via machine learning, a robust content categorization model based on the augmented training data.
Claim 14: The medium of claim 8, wherein the information, when read by the machine, further causes the machine to perform the following steps: receiving content to be categorized; and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data.
Claim 14: The medium of claim 8, wherein the information, when read by the machine, further causes the machine to perform the following steps: receiving content to be categorized; and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data.
Claim 15: A system, comprising: a training data augmenter implemented by a processor and configured for receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features, and generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels, a data sample in the supervised training data with a label and a new data sample from the weakly labeled training data with the same label have varying characteristics; and an augmented data-based model training engine implemented by a processor and configured for obtaining augmented training data based on the supervised training data and the weakly labeled training data, and training, via machine learning, a robust content categorization model based on the augmented training data.
Claim 15: A system, comprising: a training data augmenter implemented by a processor and configured for receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features, generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes data samples each of which is assigned with one of the plurality of labels via consistent self-training so that a data sample in the supervised training data with a label and a data sample from the weakly labeled training data with the same label have varying characteristics; and an augmented data-based model training engine implemented by a processor and configured obtaining augmented training data based on the supervised training data and the weakly labeled training data, and training, via machine learning, a robust content categorization model based on the augmented training data.
Claim 20: The system of claim 15, further comprising a content categorization engine implemented by a processor and configured for: receiving content to be categorized; and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data.
Claim 20: The system of claim 15, further comprising a content categorization engine implemented by a processor and configured for: receiving content to be categorized; and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data.
Claim Rejections - 35 USC § 101
Claims 8-14 are further rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because the broadest reasonable interpretation of the "machine -readable medium" encompasses signals per se. The specification discloses that “machine-readable medium may take many forms, including but not limited to, a tangible storage medium, a carrier wave medium or physical transmission medium.” (see Para 62, Lines 1-4). A claim whose BRI covers both statutory and non-statutory embodiments embraces subject matter that is not eligible for patent protection and therefore is directed to non-statutory subject matter. See MPEP 2106.03(II). It is suggested that claim 8 be amended to recite a “non-transitory” computer readable medium to overcome this rejection.
Accordingly, claims 9-14, which are dependent on claim 8, are not directed to statutory subject matter, nor do they add more to being directed to statutory subject matter on their own.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
According to the first part of the analysis, in the instant case, claims 1-7 are directed to a method, and claims 15-20 are directed to a system. Thus, each of the claims falls within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter).
Regarding Claim 1:
Step 2A Prong 1:
generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, wherein the weakly labeled training data includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels, a data sample in the supervised training data with a label and a new data sample from the weakly labeled training data with the same label have varying characteristics; (This step for generating weakly labeled training data based on the supervised training data and the unlabeled data clusters is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
Step 2A Prong 2: The judicial exception is not integrated into a practical application
Additional elements:
A method, comprising: receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; (This step is directed to receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
obtaining augmented training data based on the supervised training data and the weakly labeled training data; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
and training, via machine learning, a robust content categorization model based on the augmented training data. (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra solution activity in combination of mere instructions to perform generic computer functions that are implemented to perform the disclosed abstract idea above.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
A method, comprising: receiving supervised training data and unlabeled data clusters, wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
obtaining augmented training data based on the supervised training data and the weakly labeled training data; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
and training, via machine learning, a robust content categorization model based on the augmented training data. (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed in combination of generic computer function/instructions that are implemented to perform the disclosed abstract idea above.
Regarding Claim 2:
Incorporates the rejection of Claim 1
Step 2A Prong 1: The claim does not recite additional abstract ideas.
Step 2A Prong 2: The judicial exception is not integrated into a practical application
Additional elements:
The method of claim 1, wherein the generating the weakly labeled training data comprises: accessing the unlabeled data clusters; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
and training, via machine learning, a generative augmentation model based on the unlabeled data clusters, wherein the generative augmentation model learns variations exhibited in each of the unlabeled data clusters. (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra solution activity in combination of mere instructions to perform generic computer functions that are implemented to perform the disclosed abstract idea above.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
The method of claim 1, wherein the generating the weakly labeled training data comprises: accessing the unlabeled data clusters; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
and training, via machine learning, a generative augmentation model based on the unlabeled data clusters, wherein the generative augmentation model learns variations exhibited in each of the unlabeled data clusters. (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed in combination of generic computer function/instructions that are implemented to perform the disclosed abstract idea above.
Regarding Claim 3:
Incorporates the rejection of claim 2.
Step 2A Prong 1:
The method of claim 2, further comprising: with respect to each of the data samples in the supervised training data, generating, using the generative augmentation model, one or more new data samples with a label of the data sample assigned to each of the one or more new data samples; (This step for generating one or more new data samples with a label of the data sample is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
and creating the weakly labeled training data based on the new data samples with labels assigned thereto. (This step for creating weakly labeled training data based on the data samples with labels is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
Step 2A Prong 2 and 2B: The claim does not recite any additional elements.
Regarding Claim 4:
Incorporates the rejection of claim 2.
Step 2A Prong 1:
and generating, based on the pairs of unlabeled data samples generated for the unlabeled data clusters, training data for the machine learning. (This step for generating training data for the machine learning is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
Step 2A Prong 2: The judicial exception is not integrated into a practical application
Additional elements:
The method of claim 2, wherein the training the generative augmentation model comprises: obtaining, with respect to each of the unlabeled data clusters, pairs of unlabeled data samples with a first unlabeled data sample and a second unlabeled data sample from the unlabeled data cluster; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra solution activity that are implemented to perform the disclosed abstract idea above.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
The method of claim 2, wherein the training the generative augmentation model comprises: obtaining, with respect to each of the unlabeled data clusters, pairs of unlabeled data samples with a first unlabeled data sample and a second unlabeled data sample from the unlabeled data cluster; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed that are implemented to perform the disclosed abstract idea above.
Regarding Claim 5:
Incorporates the rejection of claim 4.
Step 2A Prong 1:
(This step for generating the second data sample in the pair via the perturbation function is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
Step 2A Prong 2 and 2B: The judicial exception is not integrated into a practical application. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
The method of claim 4, further comprising: training, using the training data comprising the pairs, the generative augmentation model to learn a perturbation function so that, given the first data sample in a pair, (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
the generative augmentation model is used to (This step is adding the word “apply it” (or an equivalent) with the judicial exception, or merely applying a generic classifier as a tool to perform the abstract idea (i.e., generating – see MPEP 2105.05(f)).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere instructions to perform generic computer functions in combination of applying a generic classifier that are implemented to perform the disclosed abstract idea above.
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed in combination of applying a generic classifier that are implemented to perform the disclosed abstract idea above.
Regarding Claim 6:
Incorporates the rejection of claim 5.
Step 2A Prong 1:
assigning the label associated with the data sample from the supervised training data to the next new data sample; (This step for assigning the label associated with the data sample is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. judgement).)
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
Additional elements:
The method of claim 5, wherein the generating the one or more new data samples with assigned labels comprises: obtaining the label associated with the data sample from the supervised training data; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
proving the data sample to the generative augmentation model; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
obtaining, from the generative augmentation model, next new data sample generated based on the perturbation function; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
and repeating the steps of providing, obtaining, and assigning for the one or more times to obtain the one or more new data samples with assigned label. (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra solution activity that are implemented to perform the disclosed abstract idea above.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
The method of claim 5, wherein the generating the one or more new data samples with assigned labels comprises: obtaining the label associated with the data sample from the supervised training data; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
proving the data sample to the generative augmentation model; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
obtaining, from the generative augmentation model, next new data sample generated based on the perturbation function; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
and repeating the steps of providing, obtaining, and assigning for the one or more times to obtain the one or more new data samples with assigned label. (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed that are implemented to perform the disclosed abstract idea above.
Regarding Claim 7:
Incorporates the rejection of claim 1.
Step 2A Prong 1:
and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data. (This step for classifying content based on the robust content categorization model is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. judgement).)
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
Additional elements:
The method of claim 1, further comprising: receiving content to be categorized; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra solution activity that are implemented to perform the disclosed abstract idea above.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
The method of claim 1, further comprising: receiving content to be categorized; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed that are implemented to perform the disclosed abstract idea above.
Regarding Claim 8
Claim 8 recites a product claim having similar limitations to the method claim of claim 1. Therefore, claim 8 is rejected for the same reasons as disclosed for claim 1 above. The additional elements of claim 8 are analyzed below:
Step 2A Prong 1: Please see Step 2A Prong 1 analysis of claim 1.
Step 2A Prong 2 and 2B: The judicial exception is not integrated into a practical application. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
A machine-readable medium having information recorded thereon, wherein the information, when read by machine, causes the machine to perform the following steps: (This step is adding the word “apply it” (or an equivalent) with the judicial exception, or merely applying a generic classifier as a tool to perform the abstract idea (i.e., generating – see MPEP 2105.05(f)).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are merely applying it with a generic classifier that are implemented to perform the disclosed abstract idea above.
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are merely applying it with a generic classifier that are implemented to perform the disclosed abstract idea above.
Regarding Claims 9-14:
Incorporates the rejection of claim 8.
Claims 9-14 recite a product claim having similar limitations to the method claims of claims 2-7. Therefore, claims 9-14 are rejected for the same reasons as disclosed for claims 2-7 above.
Regarding Claim 15:
Claim 15 recites a system claim having similar limitations to the method claim of claim 1. Therefore, claim 15 is rejected for the same reasons as disclosed for claim 1 above. The additional elements of claim 15 are analyzed below:
Step 2A Prong 1:
Step 2A Prong 2 and 2B: The judicial exception is not integrated into a practical application. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional element:
A system, comprising: a training data augmenter implemented by a processor and configured for (This step is adding the word “apply it” (or an equivalent) with the judicial exception, or merely applying a generic classifier as a tool to perform the abstract idea (i.e., augmenting/receiving – see MPEP 2105.05(f)).)
and an augmented data-based model training engine implemented by a processor and configured (This step is adding the word “apply it” (or an equivalent) with the judicial exception, or merely applying a generic classifier as a tool to perform the abstract idea (i.e., augmenting/obtaining – see MPEP 2105.05(f)).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are merely applying it with a generic classifier that are implemented to perform the disclosed abstract idea above.
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are merely applying it with a generic classifier that are implemented to perform the disclosed abstract idea above.
Regarding claim 16:
Incorporates the rejection of claim 15.
Step 2A Prong 1:
with respect to each of the data samples in the supervised training data, generating, (This step for generating one or more new data samples with a label of the data sample is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
and creating the weakly labeled training data based on the new data samples with labels assigned thereto. (This step for creating weakly labeled training data based on the data samples with labels is practically performable in the human mind and is understood to be recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
The system of claim 15, wherein the generating the weakly labeled training data comprises: accessing the unlabeled data clusters; (This step is directed to transmitting or receiving information, which is understood to be insignificant extra-solution activity and data gathering. See MPEP 2106.05(g).)
training, via machine learning, a generative augmentation model based on the unlabeled data clusters, wherein the generative augmentation model learns variations exhibited in each of the unlabeled data clusters; (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
(This step is adding the word “apply it” (or an equivalent) with the judicial exception, or merely applying a generic classifier as a tool to perform the abstract idea (i.e., generating – see MPEP 2105.05(f)).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra solution activity in combination of mere instructions to perform generic computer functions and applying a generic classifier that are implemented to perform the disclosed abstract idea above.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
The system of claim 15, wherein the generating the weakly labeled training data comprises: accessing the unlabeled data clusters; (This step is directed to transmitting or receiving information and is a well understood, routine and conventional activity as identified by the court (MPEP 2106.0f(d)(II)(i)).)
training, via machine learning, a generative augmentation model based on the unlabeled data clusters, wherein the generative augmentation model learns variations exhibited in each of the unlabeled data clusters; (Training a model (e.g. categorization model) is understood as mere instructions to implement an abstract idea (e.g., categorize content) on a computer – see MPEP 2106.05(f).)
(This step is adding the word “apply it” (or an equivalent) with the judicial exception, or merely applying a generic classifier as a tool to perform the abstract idea (i.e., generating – see MPEP 2105.05(f)).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed in combination of generic computer function/instructions and applying a generic classifier that are implemented to perform the disclosed abstract idea above.
Regarding Claim 17-20:
Claims 17-20 recite system claims having similar limitations to the method claims of claims 4-7. Therefore, claims 17-20 are rejected for the same reasons as disclosed for claims 4-7 above.
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.
Claim(s) 1-3, 7-10, 14-16, and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Taccari et al. (US 20230116417 A1, hereinafter Taccari) in view of Li et al. ("A Novel Generative Model With Bounded-GAN for Reliability Classification of Gear Safety", hereinafter Li).
Regarding Claim 1
Taccari discloses:
A method, comprising: ([Fig. 1A] discloses the method further discloses below.)
receiving supervised training data and unlabeled data clusters, ([Para 16; Fig. 1A; Para 27] discloses receiving labeled data (i.e. supervised training data) and unlabeled data items (i.e. unlabeled data). Figure 1A discloses this visually at reference numbers 102 and 104. Paragraph 27 discloses how the unlabeled data can be in sets or subsets (i.e. clusters))
wherein the supervised training data include data samples each of which has a label from a plurality of labels and each of the unlabeled data clusters includes multiple unlabeled data samples with varying features; ([Para 17; Para 21; Fig 1A; Para 27; Para 45] Paragraph 17 discloses the labeled data (i.e. supervised training data) with labels that come from a list of classifications (i.e. plurality of labels). Figure 1A further shows how the labels can include a plurality by showcasing “Dog” or “Cat” at reference number 101. Paragraph 21 discloses how the unlabeled data images can be represented with different or a variety of parameters (i.e. varying features) and can be different from set to set. Figure 1A additionally discloses images of the cat in different poses (i.e. varying features) at reference number 103. Paragraph 27 discloses the subsets (i.e. clusters) being partitioned so that the same image is represented in the group. This means that each subset (i.e. clusters) have varying similar features. Paragraph 45 discloses that in subsets of unlabeled data, the images could also have different features while keeping the general image as the same object (i.e. varying).)
generating weakly labeled training data based on the supervised training data and the unlabeled data clusters, ([Para 46; Para 47] discloses mapping a group of unlabeled data (i.e. unlabeled data clusters) with classifications and labels associated with the input labeled data (i.e. supervised training data). Paragraph 47 further discloses applying pseudo-labels based on the greatest accuracy to the first set of data.)
([Para 21] discloses how parameters (i.e. characteristics) can be different or in variety in accordance to the labeled data. It can be inferred that the new data sample from the weakly labeled data with the same label as labeled data would display a variety of parameters as well.)
obtaining augmented training data based on the supervised training data and the weakly labeled training data; ([Para 28] discloses training a model based on both the labeled data and the pseudo labeled data from the unlabeled data. The semi-supervised machine learning disclosed in this paragraph uses both data sets, which in known to be augmented data.)
and training, via machine learning, a robust content categorization model based on the augmented training data. ([Para 28; Para 59] discloses training a model based on the set of labeled data and the set of pseudo labeled data (i.e. augmented data). Paragraph 39 discloses how the trained model may be used to recognize data, or to classify images (i.e. categorization model).)
Taccari does not explicitly disclose:
wherein the weakly labeled training data includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels,
However, Li discloses in the same field of endeavor: wherein the weakly labeled training data includes new data samples each of which is generated via generative augmentation with assigned one of the plurality of labels, a data sample in the supervised training data with a label and a new data sample from the weakly labeled training data with the same label have varying characteristics; ([Page 8773 Left Co; Page 8772 Abstract; Page 8777 Right Col Para 4; Page 8777 Right Col Algorithm 1] Li discloses generating gear data and using a MCL scheme to label the generated data in view of gear data. Page 8772 discloses in the abstract that the gear data is generated based on existing gear data. Page 8777, right column, paragraph 4, starting with “For each class” discloses integrating the synthetic data with labels into the real data which is used to train classifiers. The integration of the synthetic data into the real data is augmentation. Additionally, on the same page, right column, Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. Parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data. The algorithm ends by integrating the instances into both the real data and synthetic data to add noise, effectively creating augmented data.)
Taccari and Li are both analogous art to the present invention because both are from the same field of endeavor directed to training models using data augmentation for classification.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the method for training using augmented data disclosed by Taccari with the method of training classifier models with generated synthetic data which is labeled disclosed by Li. One of ordinary skill in the art would have been motivated to make this in order to create a more robust categorization model by improving the training samples through augmentation, in addition to creating a more robust augmentation data set by adding generated data.
Regarding Claim 2
Taccari discloses in view of Li: The method of claim 1.
Taccari further discloses:
wherein the generating the weakly labeled training data comprises: accessing the unlabeled data clusters; ([Fig 1A; Para 16; Para 27] Paragraph 16 and Figure 1A discloses receiving unlabeled data. Paragraph 27 discloses the unlabeled data set being in sets and subsets (i.e. clusters). Receiving the data allows access to the data.)
and training, via machine learning, a generative augmentation model based on the unlabeled data clusters, wherein the generative augmentation model learns variations exhibited in each of the unlabeled data clusters. ([Para 45] discloses a mapping function that learns from parameters of unlabeled data partitioned into groups by an unspecified label. The model learns variations from different groups of unlabeled data and generates pseudo-labels for the data that corresponds to the groups.)
Regarding Claim 3
Taccari discloses in view of Li: The method of claim 2, as disclosed in the rejection of claim 2 above.
Li further discloses:
further comprising: with respect to each of the data samples in the supervised training data, generating, using the generative augmentation model, one or more new data samples with a label of the data sample assigned to each of the one or more new data samples; ([Page 8777 Section B; Page 8777 Right Col Algorithm 1] Li discloses generating gear data and using a MCL scheme to label the generated data in view of gear data. Page 8777, Section B discloses how the newly generated measures distance from the real data to evaluate what pseudo-labels to apply to the generated data. The distances are evaluated and sorted, and the data that is within α is integrated into the real data with pseudo-labels. Additionally, on the same page, right column, Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. Parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data. The algorithm ends by integrating the instances into both the real data and synthetic data to add noise, effectively creating augmented data.)
and creating the weakly labeled training data based on the new data samples with labels assigned thereto. ([Page 8777 Right Col Algorithm 1] Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. The distances are sorted from smallest to largest, then parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data. Data within α distance from the real data will be tagged with pseudo-labels. The algorithm ends by integrating the instances into both the real data and synthetic data to add noise, effectively creating augmented data.)
Regarding Claim 7
Taccari in view of Li discloses: The method of claim 1,
Li further discloses:
further comprising: receiving content to be categorized; ([Page 8777 right col para 1] discloses using gear parameter data to test a classification model trained on the augmented data as disclosed in claim 1.)
and classifying the content based on the robust content categorization model trained based on the augmented training data including both the supervised training data and the weakly labeled training data. ([Page 8777 right col para 1] discloses classifying the gear data in an experiment, using classifiers that are trained on the augmented data and the labeled data. Algorithm 1 discloses combining the real data with the synthetic data to train classifiers (Page 8777 left col para 5).)
Regarding Claim 8:
Claim 8 recites a product that performs the method as described in claim 1. Therefore, claim 8 is rejected under the same reasons mentioned for claim 1. The additional elements of claim 8 are disclosed below by Taccari:
Additional elements:
A machine-readable medium having information recorded thereon, wherein the information, when read by machine, causes the machine to perform the following steps: ([Para 112] discloses using a computer-readable medium which has instructions that performs the method of the invention.)
Regarding Claims 9-10:
Claims 9-10 recite a product that performs the method as described in claims 2-3. Therefore, claims 9-10 are rejected under the same reasons mentioned for claims 2-3.
Regarding Claim 14:
Claim 14 recites a product that performs the method as described in claim 7. Therefore, claim 14 is rejected under the same reasons mentioned for claim 7. The additional elements of claim 14 are disclosed below by Taccari:
Additional elements:
The medium of claim 8, wherein the information, when read by the machine, further causes the machine to perform the following steps: ([Para 112] discloses using a computer-readable medium which has instructions that performs the method of the invention.)
Regarding Claim 15:
Claim 15 recites a system that performs the method as described in claim 1. Therefore, claim 15 is rejected under the same reasons mentioned for claim 1.
Regarding Claim 16:
Taccari discloses in view of Li: The system of claim 15.
Taccari further discloses:
wherein the generating the weakly labeled training data comprises: accessing the unlabeled data clusters; ([Fig 1A; Para 16; Para 27] Paragraph 16 and Figure 1A discloses receiving unlabeled data. Paragraph 27 discloses the unlabeled data set being in sets and subsets (i.e. clusters). Receiving the data allows access to the data.)
training, via machine learning, a generative augmentation model based on the unlabeled data clusters, wherein the generative augmentation model learns variations exhibited in each of the unlabeled data clusters; ([Para 45] discloses a mapping function that learns from parameters of unlabeled data partitioned into groups by an unspecified label. The model learns variations from different groups of unlabeled data and generates pseudo-labels for the data that corresponds to the groups.)
Taccari does not disclose:
with respect to each of the data samples in the supervised training data, generating, using the generative augmentation model, one or more new data samples with a label of the data sample assigned to each of the one or more new data samples;
and creating the weakly labeled training data based on the new data samples with labels assigned thereto.
However, Li discloses in the same field of endeavor: with respect to each of the data samples in the supervised training data, generating, using the generative augmentation model, one or more new data samples with a label of the data sample assigned to each of the one or more new data samples;
([Page 8777 Section B; Page 8777 Right Col Algorithm 1] Li discloses generating gear data and using a MCL scheme to label the generated data in view of gear data. Page 8777, Section B discloses how the newly generated measures distance from the real data to evaluate what pseudo-labels to apply to the generated data. The distances are evaluated and sorted, and the data that is within α is integrated into the real data with pseudo-labels. Additionally, on the same page, right column, Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. Parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data. The algorithm ends by integrating the instances into both the real data and synthetic data to add noise, effectively creating augmented data.)
and creating the weakly labeled training data based on the new data samples with labels assigned thereto. ([Page 8777 Right Col Algorithm 1] Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. The distances are sorted from smallest to largest, then parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data. Data within α distance from the real data will be tagged with pseudo-labels. The algorithm ends by integrating the instances into both the real data and synthetic data to add noise, effectively creating augmented data.).
Regarding Claim 20:
Claim 20 recites a system that performs the method as described in claim 7. Therefore, claim 20 is rejected under the same reasons mentioned for claim 7.
Claim(s) 4-6, 11-13, 17-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Taccari et al. (US 20230116417 A1, hereinafter Taccari) in view of Li et al. ("A Novel Generative Model With Bounded-GAN for Reliability Classification of Gear Safety", hereinafter Li) in view of Fan et al. (“When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning”, hereinafter Fan).
Regarding Claim 4
Taccari discloses in view of Li: The method of claim 2, as disclosed in the rejection of claim 2 above.
Taccari in view of Li does not disclose: wherein the training the generative augmentation model comprises: obtaining, with respect to each of the unlabeled data clusters, pairs of unlabeled data samples with a first unlabeled data sample and a second unlabeled data sample from the unlabeled data cluster;
and generating, based on the pairs of unlabeled data samples generated for the unlabeled data clusters, training data for the machine learning.
However, Fan discloses in the same field of endeavor: wherein the training the generative augmentation model comprises: obtaining, with respect to each of the unlabeled data clusters, pairs of unlabeled data samples with a first unlabeled data sample and a second unlabeled data sample from the unlabeled data cluster; ([Page 2 Section 2 Para 2; Page 5 Para 1; Page 5 Para 5] Page 2 discloses background and related work to the art, which discusses SimCLR, that Fan builds off of. While not being pertinent to the invention directly, Fan uses the unlabeled source data set and obtains pairs of unlabeled data from them. Page 5 discloses using pairs of unlabeled data T1(x) and T2(x), which is derived from the SimCLR method disclosed in the background, to obtain paired perturbed data (Page 5 Para 5).)
and generating, based on the pairs of unlabeled data samples generated for the unlabeled data clusters, training data for the machine learning. ([Page 6 Para 4] discloses generating pseudo labels on the generated data using ClusterFit to map cluster indexes of data to a pseudo-label to make a pseudo-labeled data set.)
Taccari, Li, and Fan are analogous art to the present invention because they are from the same field of endeavor directed to training models using unlabeled and labeled data through data augmentation and semi-supervised learning.
It would have been obvious for one of the ordinary skill in the art before the effective filing date of the claimed invention to have combined the data augmentation methods proposed by Taccari with the labeling method with data augmentation proposed by Li with the contrastive learning method proposed by Fan. One would be motivated to combine these systems to create a robust data augmentation system, including various methods to create complete training data, including transformations on real data.
Regarding Claim 5
Taccari discloses in view of Li and Fan: The method of claim 4, as disclosed in the rejection of claim 4 above.
Fan discloses further:
training, using the training data comprising the pairs, the generative augmentation model to learn a perturbation function so that, given the first data sample in a pair, the generative augmentation model is used to generate the second data sample in the pair via the perturbation function, wherein the second data sample generated corresponds to a varying version of the first data sample in the pair. ([Page 5 Para 4-5; Page 6 Para 1-2] using a adversarial perturb generator to create a view of the input data sample x. Fan discloses the use of perturbed data to obtain more contrasting data for an improved data set. Page 6, paragraph 1-2 discloses perturbation tolerance used during training that focuses on contrastive loss. Page 6 Paragraph 2 further discloses that the input data is unlabeled.)
Regarding Claim 6
Taccari discloses in view of Li and Fan: The method of claim 5, as disclosed in the rejection of claim 5 above.
Li further discloses:
wherein the generating the one or more new data samples with assigned labels comprises: obtaining the label associated with the data sample from the supervised training data; (Li [Page 8777 Algorithm 1] discloses using real data with labels DO as input to label generated data.)
proving the data sample to the generative augmentation model; ([Page 8773 Left Col] discloses providing real gear data to the GAN to generate new data samples.)
assigning the label associated with the data sample from the supervised training data to the next new data sample; ([Page 8777 right col Algorithm 1] Li discloses an algorithm for the labeling scheme for the generated data. The algorithm takes in an input data set of real data “D0” which has labels, and generated date “DB” without labels and outputs synthetic data “DF” with labels. The algorithm calculates the distance of the generated data with respect to the real data and applies labels that are closest to the distance calculated. Parameter α is used to approximate which group the labels will be applied to and control the instances of newly labeled data.)
and repeating the steps of providing, obtaining, and assigning for the one or more times to obtain the one or more new data samples with assigned label. ([Page 8777 right col Algorithm 1] discloses an algorithm for labeling generated data. The algorithm loops until the real data is integrated with the synthetic data for augmented data, hence repeating.)
Li does not disclose:
However, Fan discloses in the same field of endeavor: obtaining, from the generative augmentation model, next new data sample generated based on the perturbation function; ([Page 5 Para 4-5; Page 6 Para 7] discloses adversarial perturbation to generate perturbed images. Page 6, paragraph 7 discloses classification on perturbed images. This is a step in this limitation for assigning labels to the generated samples.)
Regarding Claims 11-13
Claims 11-13 recite a product that performs the method as described in claims 4-6. Therefore, claims 11-13 are rejected under the same reasons mentioned for claims 4-6.
Regarding Claims 17-19
Claims 17-19 recite a system that performs the method as described in claims 4-6. Therefore, claims 17-19 are rejected under the same reasons mentioned for claims 4-6.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SUMAIR R CHOWDHURY whose telephone number is (571)270-0523. The examiner can normally be reached Monday-Friday 8:30am - 6pm.
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, ABDULLAH AL KAWSAR can be reached at (571) 270-3169. 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.
/SUMAIR RASHED CHOWDHURY/ Examiner, Art Unit 2127
/ABDULLAH AL KAWSAR/Supervisory Patent Examiner, Art Unit 2127
Prosecution Insights