Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
This Office Action is in response to the communication filed on 24 Jul 2023.
Claims 1-11 are being considered on the merits.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 06 Apr 2026 has been considered. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, initialed and dated copies of Applicant's IDS form 1499 is attached to the instant Office action.
Drawings
The drawings filed on 24 Jul 2023 are accepted.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitations use a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations and their associated disclosures are:
first augmentation unit configured to generate a plurality of first modified vectors – In claims 1, 2 and 11 (Specification paras 6-7 and 16 provide some functional description)
second augmentation unit configured to generate a plurality of second modified vectors – In claims 1, 2 and 11 (Specification paras 6-7 and 16 provide some functional description)
loss calculation unit configured to calculate a loss – In claims 1, 7, 8, and 11 (Specification paras 0012-13, 16, and 38-39 provide some functional description)
optimization unit – In claims 1 and 11 (Specification paras 6, 16 and 39 provide some functional description)
first dimension conversion unit configured to convert dimensions – In claims 9 and 11 (Specification para 16 provides some functional description)
second dimension conversion unit configured to convert dimensions – In claims 9 and 11 (Specification para 16 provides some functional description)
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recite sufficient structure to perform the claimed function so as to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 2, 7-9, and 11 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
In claims 1, 2, and 11, claim limitation “first augmentation unit configured to generate a plurality of first modified vectors” invokes 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. However, the written description fails to disclose the corresponding structure, material, or acts for performing the entire claimed function and to clearly link the structure, material, or acts to the function. No clear association between the structure and the function can be found in the specification. As a result, the specification is unclear as to whether the structure of the first augmentation unit is software or hardware, or some other structure altogether. Therefore, the claim is indefinite and is rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
In claims 1, 2, and 11, claim limitation “second augmentation unit configured to generate a plurality of first modified vectors” invokes 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. However, the written description fails to disclose the corresponding structure, material, or acts for performing the entire claimed function and to clearly link the structure, material, or acts to the function. No clear association between the structure and the function can be found in the specification. As a result, the specification is unclear as to whether the structure of the second augmentation unit is software or hardware, or some other structure altogether. Therefore, the claim is indefinite and is rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
In claims 1, 7, 8, and 11, claim limitation “loss calculation unit configured to calculate a loss” invokes 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. However, the written description fails to disclose the corresponding structure, material, or acts for performing the entire claimed function and to clearly link the structure, material, or acts to the function. No clear association between the structure and the function can be found in the specification. As a result, the specification is unclear as to whether the structure of the loss calculation unit is software or hardware, or some other structure altogether. Therefore, the claim is indefinite and is rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
In claims 1 and 11, claim limitation “optimization unit configured to optimize parameters” invokes 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. However, the written description fails to disclose the corresponding structure, material, or acts for performing the entire claimed function and to clearly link the structure, material, or acts to the function. No clear association between the structure and the function can be found in the specification. As a result, the specification is unclear as to whether the structure of the optimization unit is software or hardware, or some other structure altogether. Therefore, the claim is indefinite and is rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
In claims 9 and 11, claim limitation “first dimension conversion unit configured to convert dimensions” invokes 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. However, the written description fails to disclose the corresponding structure, material, or acts for performing the entire claimed function and to clearly link the structure, material, or acts to the function. No clear association between the structure and the function can be found in the specification. As a result, the specification is unclear as to whether the structure of the first dimension conversion unit is software or hardware, or some other structure altogether. Therefore, the claim is indefinite and is rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
In claims 9 and 11, claim limitation “second dimension conversion unit configured to convert dimensions” invokes 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. However, the written description fails to disclose the corresponding structure, material, or acts for performing the entire claimed function and to clearly link the structure, material, or acts to the function. No clear association between the structure and the function can be found in the specification. As a result, the specification is unclear as to whether the structure of the second dimension conversion unit is software or hardware, or some other structure altogether. Therefore, the claim is indefinite and is rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
Applicant may:
(a) Amend the claim so that the claim limitation will no longer be interpreted as a limitation under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph;
(b) Amend the written description of the specification such that it expressly recites what structure, material, or acts perform the entire claimed function, without introducing any new matter (35 U.S.C. 132(a)); or
(c) Amend the written description of the specification such that it clearly links the structure, material, or acts disclosed therein to the function recited in the claim, without introducing any new matter (35 U.S.C. 132(a)).
If applicant is of the opinion that the written description of the specification already implicitly or inherently discloses the corresponding structure, material, or acts and clearly links them to the function so that one of ordinary skill in the art would recognize what structure, material, or acts perform the claimed function, applicant should clarify the record by either:
(a) Amending the written description of the specification such that it expressly recites the corresponding structure, material, or acts for performing the claimed function and clearly links or associates the structure, material, or acts to the claimed function, without introducing any new matter (35 U.S.C. 132(a)); or
(b) Stating on the record what the corresponding structure, material, or acts, which are implicitly or inherently set forth in the written description of the specification, perform the claimed function. For more information, see 37 CFR 1.75(d) and MPEP §§ 608.01(o) and 2181.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1 and 10-11 are rejected under 35 U.S.C. 103 as being unpatentable over Choi, Hee Min & Hyoa Kang (US 2022/0237890 A1; hereinafter, “Choi”) in view of Wang, et. al. (arXiv:2004.05937v7 [cs.CV] 17 Jun 2021; hereinafter, “Wang”).
Regarding claim 1, Choi as modified teaches:
A network learning system comprising: a first augmentation unit configured to generate a plurality of first modified vectors by modifying a first feature vector outputted from a teacher network; a second augmentation unit configured to generate a plurality of second modified vectors by modifying a second feature vector outputted from a student network (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”; Choi, para. 0072 “The first neural network model 221 may also be referred to herein as a teacher model, and the second neural network model 222 may also be referred to herein as a student model.”; Choi para. 0073-0074 and fig. 1 teaches a first and second view generation model which takes feature data output from a first and second neural network (a teacher and student model) and calculate second embedded data from the feature data i.e. a modified feature vector); a loss calculation unit configured to calculate a loss by using the first modified vectors and the second modified vectors; (Choi, para. 0116: “A discriminative loss calculator 660 may calculate an adversarial loss between embedded data output from one neural network model among the n neural network models and embedded data output from another neural network model among the n neural network models in a similar way as represented by Equation 15 above.”) and an optimization unit (Choi, para. 0143: “ Examples of hardware components that may be used to perform the operations described in this application where appropriate include controllers, sensors, generators, drivers, memories, comparators, arithmetic logic units, adders, subtractors, multipliers, dividers, integrators, and any other electronic components configured to perform the operations described in this application.” Examiner notes Choi teaches components may be used to perform operations where appropriate) configured to optimize parameters of the student network based on the calculated loss. (Wang, sec. 8.1.2: “By simultaneously optimizing the distillation and the adversarial loss, the classifier(student) learns the true data distribution at equilibrium” Examiner notes Wang teaches optimizing parameters of the student network and Choi teaches hardware components, i.e. a unit, which may perform operations including optimization operations taught by Wang).
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Wang into Choi. Choi teaches dynamic cluster-based search and retrieval; Wang teaches A semantic search system integrates with an AI platform to provide advanced search capabilities by leveraging automatically generated ontologies and knowledge graphs. One of ordinary skill would have been motivated to combine the teachings of Wang into Choi in order to enable efficient learning of the student while using small amount of training data (Wang, sec. 5.2).
Regarding claim 10, Choi as modified teaches:
A network learning method comprising: a first augmentation step of generating a plurality of first modified vectors by modifying a first feature vector outputted from a teacher network; a second augmentation step of generating a plurality of second modified vectors by modifying a second feature vector outputted from a student network (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”; Choi, para. 0072 “The first neural network model 221 may also be referred to herein as a teacher model, and the second neural network model 222 may also be referred to herein as a student model.”; Choi para. 0073-0074 and fig. 1 teaches a first and second view generation model which takes feature data output from a first and second neural network (a teacher and student model) and calculate second embedded data from the feature data i.e. a modified feature vector); a loss calculation step of calculating a loss by using the first modified vectors and the second modified vectors (Choi, para. 0116: “A discriminative loss calculator 660 may calculate an adversarial loss between embedded data output from one neural network model among the n neural network models and embedded data output from another neural network model among the n neural network models in a similar way as represented by Equation 15 above.”); and an optimization step of optimizing parameters of the student network based on the calculated loss. (Wang, sec. 8.1.2: “By simultaneously optimizing the distillation and the adversarial loss, the classifier(student) learns the true data distribution at equilibrium” Examiner notes Wang teaches optimizing parameters of the student network based on calculated adversarial loss).
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Wang into Choi, as modified, as set forth above with respect to claim 1.
Regarding claim 11, Choi as modified teaches:
A network learning system comprising: a first augmentation unit configured to generate a plurality of first modified vectors by modifying a first feature vector outputted from a teacher network; a second augmentation unit configured to generate a plurality of second modified vectors by modifying a second feature vector outputted from a student network (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”; Choi, para. 0072 “The first neural network model 221 may also be referred to herein as a teacher model, and the second neural network model 222 may also be referred to herein as a student model.”; Choi para. 0073-0074 and fig. 1 teaches a first and second view generation model which takes feature data output from a first and second neural network (a teacher and student model) and calculate second embedded data from the feature data i.e. a modified feature vector); a first dimension conversion unit configured to convert dimensions of the first modified vectors generated in the first augmentation unit; (Choi, para. 0073: “The first embedded data may be data projected to a lower dimension than a dimension of the first backbone feature data, and the first view data may be data from which some features are dropped out.”) a second dimension conversion unit configured to convert dimensions of the second modified vectors generated in the second augmentation unit (Choi, para. 0074: “The second embedded data may be data projected to a lower dimension than a dimension of the second backbone feature data, and the second view data may be data from which some features are dropped out.”); a loss calculation unit configured to calculate a loss by using the first modified vectors and the second modified vectors the dimensions of which are converted; (Choi, para. 0116: “A discriminative loss calculator 660 may calculate an adversarial loss between embedded data output from one neural network model among the n neural network models and embedded data output from another neural network model among the n neural network models in a similar way as represented by Equation 15 above.” Examiner notes Choi teaches loss using the first and second vectors whose dimensions are converted) and an optimization unit (Choi, para. 0143: “Examples of hardware components that may be used to perform the operations described in this application where appropriate include controllers, sensors, generators, drivers, memories, comparators, arithmetic logic units, adders, subtractors, multipliers, dividers, integrators, and any other electronic components configured to perform the operations described in this application.” Examiner notes Choi teaches components, i.e. units, which may be used to perform operations where appropriate) configured to optimize parameters of the student network based on the calculated loss. (Wang, sec. 8.1.2: “By simultaneously optimizing the distillation and the adversarial loss, the classifier(student) learns the true data distribution at equilibrium” Examiner notes Wang teaches optimizing parameters of the student network and Choi teaches hardware components, i.e. a unit, which may perform operations including optimization operations taught by Wang).
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Wang into Choi, as modified, as set forth above with respect to claim 1.
Claims 2 and 6-9 are rejected under 35 U.S.C. 103 as being unpatentable over Choi in view of Wang, and further in view of Kida, Singo (US 2024/0338605 A1; hereinafter, “Kida”).
Regarding claim 2, Choi as modified teaches:
The network learning system of claim 1, wherein the first augmentation unit (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) comprises first modifiers configured to generate the plurality of first modified vectors by applying different modification methods to the first feature vector, (Kida, para. 0044-0045: “The MergeNet36 processes the average basic feature vector and the average novel feature vector by a neural network and outputs task-specific merged weight vectors ωpre and ωmeta for calculating the task-adaptive representation TAR… The backbone CNN 22 operates as a basic feature vector extractor fθ that extracts a basic feature vector for an input x and outputs a basic feature vector fθ (x) for the input x. The intermediate layer output of the backbone CNN22 in response to the input x is denoted by ac (x). The MetaCNN 32 operates as a novel feature vector extractor g that extracts a novel feature vector for the intermediate layer output aθ(x), and outputs a novel feature vector g (aθ(x)) for the intermediate layer output aθ(x).” Examiner notes Kida teaches a modifier, MergeNet, to generate modified vectors in the form of task-adaptive representation, TAR, where the modifiers are different depending on the task.).
wherein the second augmentation unit (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) comprises second modifiers configured to generate the plurality of second modified vectors by applying different modification methods to the second feature vector, and (Kida, para. 0044-0045: “The MergeNet36 processes the average basic feature vector and the average novel feature vector by a neural network and outputs task-specific merged weight vectors ωpre and ωmeta for calculating the task-adaptive representation TAR… The backbone CNN 22 operates as a basic feature vector extractor fθ that extracts a basic feature vector for an input x and outputs a basic feature vector fθ (x) for the input x. The intermediate layer output of the backbone CNN22 in response to the input x is denoted by ac (x). The MetaCNN 32 operates as a novel feature vector extractor g that extracts a novel feature vector for the intermediate layer output aθ(x), and outputs a novel feature vector g (aθ(x)) for the intermediate layer output aθ(x).” Examiner notes Kida teaches a modifier, MergeNet, to generate modified vectors in the form of task-adaptive representation, TAR where the modifiers are different depending on the task).
wherein the first modifiers and the second modifiers are configured to make pairs and apply a same modification method. (Choi, para. 0061: “Meanwhile, a student model of one or more embodiments may be trained to mimic a prediction of an on-the-fly self-supervised teacher.” Examiner notes for examination purposes only, “configured to make pairs” is interpreted as creating a pairing with each other, where Choi teaches a student and teacher model making a pair by mimicking each other).
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Kida into Choi as modified. Kida teaches a machine learning apparatus that continually learns a novel class with fewer samples than a base class. One of ordinary skill would have been motivated to combine the teachings of Kida into Choi as modified in order to reduce the learning time by pre-training a feature extractor and allowing loss to converge easily (Kida, para 0075).
Regarding claim 6, Choi as modified teaches:
The network learning system of claim 3, wherein the first modifier (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) is configured to generate a first weight vector from the first modified vector, to generate a second weight vector from the first feature vector (Kida, para. 0044: “The MergeNet36 processes the average basic feature vector and the average novel feature vector by a neural network and outputs task-specific merged weight vectors ωpre and ωmeta for calculating the task-adaptive representation TAR.” Examiner teaches a first and second weight vector, ωpre and ωmeta), to generate a final weight vector by adding up the first weight vector and the second weight vector, and to generate a final first modified vector by calculating the first feature vector and the final weight vector, and (Kida, para. 0049: “The vector sum arithmetic unit 37 calculates a vector sum of i) a product between the basic feature vector fθ(x) and the merged weight vector ωpre and ii) a product between the novel feature vector g (aθ(x)) and the merged weight vector ωmeta.”
wherein the second modifier (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) is configured to generate a first weight vector from the second modified vector, to generate a second weight vector from the second feature vector (Kida, para. 0044: “The MergeNet36 processes the average basic feature vector and the average novel feature vector by a neural network and outputs task-specific merged weight vectors ωpre and ωmeta for calculating the task-adaptive representation TAR.” Examiner teaches a first and second weight vector, ωpre and ωmeta), to generate a final weight vector by adding up the first weight vector and the second weight vector, and to generate a final second modified vector by calculating the second feature vector and the final weight vector. (Kida, para. 0049: “The vector sum arithmetic unit 37 calculates a vector sum of i) a product between the basic feature vector fθ(x) and the merged weight vector ωpre and ii) a product between the novel feature vector g (aθ(x)) and the merged weight vector ωmeta.”
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Kida into Choi, as modified, as set forth above with respect to claim 2.
Regarding claim 7, Choi as modified teaches:
The network learning system of claim 2, wherein the loss calculation unit (Choi, para. 0143: “ Examples of hardware components that may be used to perform the operations described in this application where appropriate include controllers, sensors, generators, drivers, memories, comparators, arithmetic logic units, adders, subtractors, multipliers, dividers, integrators, and any other electronic components configured to perform the operations described in this application.” Examiner notes Choi teaches components may be used to perform operations where appropriate) is configured to calculate an average of differences between the first modified vectors and the second modified vectors as a loss. (Choi, para. 0116: “A discriminative loss calculator 660 may calculate an adversarial loss between embedded data output from one neural network model among the n neural network models and embedded data output from another neural network model among the n neural network models in a similar way as represented by Equation 15 above.”)
Regarding claim 8, Choi as modified teaches:
The network learning system of claim 7, wherein the loss calculation unit is configured to further calculate an average of differences between a first modified vector having smallest modification among the first modified vectors, and the first modified vectors, as a loss. (Choi, para. 0115: “A clustering unit 640 may perform clustering on embedded data output from each of the view generation models 231, 232, and 639. A cross loss calculator 650 may calculate a cross loss in a similar way as represented in Equation 14 above. A code value used to calculate the cross loss in Equation 14 may be determined as follows. For example, the cross loss calculator 650 may determine one of a code value indicating a cluster to which embedded data output by a model having the highest recognition performance among a plurality of neural network models including a first neural network model and a second neural network model belongs, a code value indicating a cluster to which embedded data output from each model belongs, and a code value indicating a cluster to which embedded data output from a model with a least clustering loss belongs. The cross loss calculator 650 may calculate the cross loss between the determined code value and the embedded data, and use the obtained cross loss for backpropagation learning or training.” Examiner notes Choi teaches a clustering unit and determining a single cluster with the least clustering loss, thereby calculating at least an average of differences in each cluster to make such determination).
Regarding claim 9, Choi as modified teaches:
The network learning system of claim 2, further comprising: a first dimension conversion unit configured to convert dimensions of the first modified vectors generated in the first augmentation unit; and (Choi, para. 0073: “The first embedded data may be data projected to a lower dimension than a dimension of the first backbone feature data, and the first view data may be data from which some features are dropped out.”)
a second dimension conversion unit configured to convert dimensions of the second modified vectors generated in the second augmentation unit. (Choi, para. 0074: “The second embedded data may be data projected to a lower dimension than a dimension of the second backbone feature data, and the second view data may be data from which some features are dropped out.”)
Claims 3-5 are rejected under 35 U.S.C. 103 as being unpatentable over Choi in view of Wang, in view of Kida, and further in view of Sundaresan, Sairam and Souvik Kundu (US 2022/0036194 A1; hereinafter, “Sundaresan”)
Regarding claim 3, Choi as modified teaches:
The network learning system of claim 2, wherein the first modifiers (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) are configured to generate the first modified vectors by masking a part of feature data constituting the first feature vector with zero, and (Sundaresan, para. 0055: “To prune the student model ΦS while simultaneously distilling knowledge from the teacher ΦT, the ΦS's total trainable parameters is/are updated, and then a mask is used to forcibly set a fraction of these parameters to zero.”)
wherein the second modifiers (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) are configured to generate the second modified vectors by masking a part of feature data constituting the second feature vector with zero. (Sundaresan, para. 0055: “To prune the student model ΦS while simultaneously distilling knowledge from the teacher ΦT, the ΦS's total trainable parameters is/are updated, and then a mask is used to forcibly set a fraction of these parameters to zero.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Sundaresan into Choi as modified. Sundaresan teaches a techniques for optimizing artificial intelligence (AI) and/or machine learning (ML) models to reduce resource consumption while maintaining or improving AI/ML model performance. One of ordinary skill would have been motivated to combine the teachings of Sundaresan into Choi as modified in order to produce models that scale better than ML models that use existing training and optimization methods (Sundaresan, para. 0033).
Regarding claim 4, Choi as modified teaches:
The network learning system of claim 3, wherein the first modifiers and the second modifiers (Choi, para. 0056: “The first embedded data and the first view data may represent a feature vector embedded in a first embedding space of the first neural network model, and the second embedded data and the second view data may represent a feature vector embedded in a second embedding space of the second neural network model.”) are configured to mask a number of pieces of feature data determined according to pre-set masking ratios with zero, respectively. (Sundaresan, para. 0055: “To prune the student model ΦS while simultaneously distilling knowledge from the teacher ΦT, the ΦS's total trainable parameters is/are updated, and then a mask is used to forcibly set a fraction of these parameters to zero. Here, the “mask” for individual layers is a binary tensor used to make sure the trainable parameters have a fixed number of non-zeros based on a given pruning budget. The mask tensor can be both irregular or structured (regular) based on the selected/chosen type of model pruning technique. In one example, the mask tensor is an upper-triangular matrix, but may be some other suitable data structure” Examiner notes Sundaresan teaches a pre-set masking in the form of a pruning budget or a pruning technique).
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Sundaresan into Choi, as modified, as set forth above with respect to claim 3.
Regarding claim 5, Choi as modified teaches:
The network learning system of claim 4, wherein the masking ratios are configured by a user or randomly configured, and wherein positions of feature data to be masked with zero are randomly determined. (Sundaresan, para. 0055: “To prune the student model ΦS while simultaneously distilling knowledge from the teacher ΦT, the ΦS's total trainable parameters is/are updated, and then a mask is used to forcibly set a fraction of these parameters to zero. Here, the “mask” for individual layers is a binary tensor used to make sure the trainable parameters have a fixed number of non-zeros based on a given pruning budget. The mask tensor can be both irregular or structured (regular) based on the selected/chosen type of model pruning technique. In one example, the mask tensor is an upper-triangular matrix, but may be some other suitable data structure”).
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Sundaresan into Choi, as modified, as set forth above with respect to claim 3.
Prior Art
X. Liu and Z. Zhu ("Knowledge Distillation for Object Detection Based on Mutual Information," 2021 4th International Conference on Intelligent Autonomous Systems (ICoIAS), Wuhan, China, 2021, pp. 18-23, doi: 10.1109/ICoIAS53694.2021.00011.) teaches a model compression scheme for object detection algorithm based on knowledge distillation.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Sally T. Ley whose telephone number is (571)272-3406. The examiner can normally be reached Monday - Thursday, 10:00am - 6:00pm ET.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached at (571) 270-5871. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/STL/Examiner, Art Unit 2147
/ERIC NILSSON/Primary Examiner, Art Unit 2151
Prosecution Insights