DETAILED ACTION
This action is in response to the claims filed 12/14/2023 for Application 18/539,431. Claims 1-16 are currently pending.
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 .
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 12/14/2023 and 01/06/2026 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
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 1-16 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.
Regarding claims 1 and 9, the claims recite “filtering the extracted feature map” in the limitation of extracting student SAMs, however the limitation is open to the interpretation of the extracted feature map being from the teacher network. Thus, it is unclear whether the filtering of the extracted feature map is tied to the feature map of teacher network or student network. For purposes of examination, the examiner will interpret the limitations as follows:
extracting teacher sparse activation maps (SAMs) by extracting a feature map from a learning model of the teacher network based on input data and filtering the extracted feature map of the teacher network;
extracting student sparse activation maps (SAMs) by extracting a feature map from a learning model of the student network based on the input data and filtering the extracted feature map of the student network;
Claims 2-8 and 10-16 are rejected as being dependent on a rejected base claim without curing any of the deficiencies.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-16 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding claim 1,
Step 1 Analysis: Claim 1 is directed to a process, which falls within one of the four statutory categories.
Step 2A Prong 1 Analysis: Claim 1 recites, in part, The limitations of:
filtering the extracted feature map (of the teacher network) can be considered to be an evaluation in the human mind,
filtering the extracted feature map (of the student network) can be considered to be an evaluation in the human mind
These limitations as drafted, are processes that, under broadest reasonable interpretation, covers performance of the limitation in the mind or with the aid of pen and paper which falls within the “Mental Processes” grouping of abstract ideas.
The claim further recites:
extracting teacher sparse activation maps (SAMs) by extracting a feature map from a learning model of the teacher network based on input data which can be considered to be a mathematical calculation.
extracting student sparse activation maps (SAMs) by extracting a feature map from a learning model of the student network based on the input data which can be considered to be a mathematical calculation.
computing a loss function by comparing the extracted teacher sparse activation maps with the extracted student sparse activation maps which can be considered to be a mathematical calculation.
and updating the learning model of the student network based on the computed loss function which can be considered a mathematical calculation.
These limitations as drafted, are processes that, under broadest reasonable interpretation, covers mathematical calculations which falls within the “Mathematical Concepts” grouping of abstract ideas.
Accordingly, the claim recites an abstract idea.
Step 2A Prong 2 Analysis: This judicial exception is not integrated into a practical application. In particular, the claim recites the additional elements – “a computing device including one or more processors and a memory that stores one or more programs executed by the one or more processors”. Thus, these elements in the claim are recited at a high level of generality such that they amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP 2106.05(f). Additionally, the claim recites the additional elements of “from a learning model of the teacher network” and “from a learning model of the student network”. These additional elements are merely generally linked to the judicial exception. Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea.
Step 2B Analysis: The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of utilizing a computing device which includes one or more processors and memory to perform the steps of the claimed process amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Additionally, the additional elements of “a learning model of the teacher network” and “a learning model of the student network” are generally linked to the judicial exception. Even when considered in combination, these additional elements amount to generally linking the elements to the judicial exception and mere instructions to apply the exception using generic computer components which cannot provide an inventive concept. The claim is not patent eligible.
Regarding claim 2, the rejection of claim 1 is further incorporated, and further, the claim recites: extracting the feature map using a convolution layer from the learning model of the teacher network based on the input data;
extracting an activation map by filtering the extracted feature map using an activation function; and
extracting the teacher sparse activation maps by filtering the extracted activation map using a filter function. This claim recites additional mathematical steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding claim 3, the rejection of claim 2 is further incorporated, and further, the claim recites: normalizing the activation map extracted from the learning model of the teacher network to a preset range; and re-filtering the normalized activation map through the activation function. This claim recites additional mathematical steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding claim 4, the rejection of claim 1 is further incorporated, and further, the claim recites: extracting the feature map using a convolution layer in the learning model of the student network based on the input data;
extracting an activation map by filtering the extracted feature map using an activation function; and
extracting the student sparse activation map by filtering the extracted activation map using a filter function. This claim recites additional mathematical steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding claim 5, the rejection of claim 4 is further incorporated, and further, the claim recites: normalizing the activation map extracted from the learning model of the student network to a preset range; and re-filtering the normalized activation map through the activation function.
This claim recites additional mathematical steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding claim 6, the rejection of claim 1 is further incorporated, and further, the claim recites: wherein the activation function is a rectified linear unit (ReLU) function, and the filter function may be a function obtained by moving the coordinates of the ReLU function and adjusting a passband.
This claim recites additional mathematical steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding claim 7, the rejection of claim 1 is further incorporated, and further, the claim recites: wherein the loss function is a loss function obtained by comparing losses due to differences in distance between the extracted teacher sparse activation map and the extracted student sparse activation map.
This claim recites additional mental steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding claim 8, the rejection of claim 1 is further incorporated, and further the claim recites:
PNG
media_image1.png
256
624
media_image1.png
Greyscale
This claim recites additional mathematical steps in addition to the judicial exception identified in the rejection of claim 1, thus recites a judicial exception.
The claim does not include any additional elements that amount to an integration of the judicial exceptions into a practical application, nor to significantly more than the judicial exceptions. The claim is not patent eligible.
Regarding Claims 9-16, they recite features similar to claims 1-8 and are rejected for at least the same reasons therein.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 4, 7-10, 12, and 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Han et al. ("US 20260004138 A1" which claims foreign priority to CN202211075557.2 filed 09/05/2022, hereinafter "Han") in view of Heo et al. ("Knowledge Transfer via Distillation of Activation Boundaries Formed by Hidden Neurons", cited by Applicant in the IDS filed 01/06/2026, hereinafter "Heo").
Regarding claim 1, Han teaches A method for learning activated neurons responses transfer using sparse activation maps (SAMs) in knowledge distillation performed on a computing device including one or more processors and a memory that stores one or more programs executed by the one or more processors (¶0117-¶0119, “the computer software program may be installed in a memory of a computing device and executed by one or more processors so as to implement corresponding functions.”), the method comprising:
extracting teacher sparse activation maps (SAMs) by extracting a feature map from a learning model of the teacher network based on input data (“At step S120, each sample image in the sample image dataset is input into the first defect detection model and the second defect detection model, respectively, and a first feature map output by a target convolutional layer in the first defect detection model and a second feature map output by a corresponding target convolutional layer in the second defect detection model are extracted” [¶0058; first detection model corresponds to the teacher as noted in ¶0059])
extracting student sparse activation maps (SAMs) by extracting a feature map from a learning model of the student network based on the input data (“At step S120, each sample image in the sample image dataset is input into the first defect detection model and the second defect detection model, respectively, and a first feature map output by a target convolutional layer in the first defect detection model and a second feature map output by a corresponding target convolutional layer in the second defect detection model are extracted” [¶0058; second detection model corresponds to the student as noted in ¶0059])
computing a loss function by comparing the extracted teacher sparse activation maps with the extracted student sparse activation maps (“At step S130, distances between corresponding feature vectors in the first feature map and the second feature map are calculated, corresponding elements in the segmentation labeling factor matrix are used to correct the distances, so as to obtain corrected distances between corresponding feature vectors in the first feature map and the second feature map, and a sum of the corrected distances between all the feature vectors in the first feature map and the second feature map is calculated as the first loss function.” [¶0066]); and
updating the learning model of the student network based on the computed loss function. (“In the present step, based on the first loss function obtained in the preceding steps, the iterative training can be performed on the second defect detection model on the basis of minimizing the first loss function, parameters of the second defect detection model are iteratively updated under specified learning rate and batch size conditions, thus ultimately yielding the proper distilled second defect detection model” [¶0077])
However Han fails to explicitly teach and filtering the extracted feature map (of the teacher and student networks);
Heo teaches filtering the extracted feature map (of the teacher and student networks); (“To transfer the activation boundaries accurately, our idea is to amplify the negligible transfer loss at the region near the activation boundaries. To amplify the loss, we define an element-wise activation indicator function that expresses
PNG
media_image2.png
68
336
media_image2.png
Greyscale
” [bottom right col of pg. 2 – top left of pg. 3; note: equation (2) acts similarly to the “filtering function” as it separates activated neurons from deactivated neurons of the student and teacher networks.])
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Han’s teachings by implementing the filtering function of Heo. One would have been motivated to make this modification in order to transfer the activation boundaries accurately between the teacher and student networks. [pg. 3, left col, ¶1, Heo]
Regarding claim 2, Han/Heo teaches The method of claim 1, wherein the extracting of the teacher sparse activation maps further comprises:
Han teaches extracting the feature map using a convolution layer from the learning model of the teacher network based on the input data (“At step S120, each sample image in the sample image dataset is input into the first defect detection model and the second defect detection model, respectively, and a first feature map output by a target convolutional layer in the first defect detection model and a second feature map output by a corresponding target convolutional layer in the second defect detection model are extracted” [¶0058]);
Heo teaches extracting an activation map by filtering the extracted feature map using an activation function (“For convenience, we assume that the teacher and the student have hidden layers of the same size. In order to describe the activation of neurons, T (I) and S(I) are defined as the values before the activation functions. Here, we consider the element-wise ReLU σ(x) = max(0,x) as the activation function.” [pg. 2, right co, ¶2]); and
extracting the teacher sparse activation maps by filtering the extracted activation map using a filter function. (“To amplify the loss, we define an element-wise activation indicator function that expresses
PNG
media_image2.png
68
336
media_image2.png
Greyscale
” [bottom right col of pg. 2 – top left of pg. 3; note: eq (2) corresponds to a “filter function”])
Same motivation to combine the teachings of Han/Heo as claim 1.
Regarding claim 4, Han/Heo teaches The method of claim 1, wherein the extracting of the student sparse activation maps further comprises:
Han teaches extracting the feature map using a convolution layer from the learning model of the student network based on the input data (“At step S120, each sample image in the sample image dataset is input into the first defect detection model and the second defect detection model, respectively, and a first feature map output by a target convolutional layer in the first defect detection model and a second feature map output by a corresponding target convolutional layer in the second defect detection model are extracted” [¶0058]);
Heo teaches extracting an activation map by filtering the extracted feature map using an activation function (“For convenience, we assume that the teacher and the student have hidden layers of the same size. In order to describe the activation of neurons, T (I) and S(I) are defined as the values before the activation functions. Here, we consider the element-wise ReLU σ(x) = max(0,x) as the activation function.” [pg. 2, right co, ¶2]); and
extracting the student sparse activation maps by filtering the extracted activation map using a filter function. (“To amplify the loss, we define an element-wise activation indicator function that expresses
PNG
media_image2.png
68
336
media_image2.png
Greyscale
” [bottom right col of pg. 2 – top left of pg. 3; note: eq (2) corresponds to a “filter function”])
Same motivation to combine the teachings of Han/Heo as claim 1.
Regarding claim 7, Han/Heo teaches The method of claim 1, Han further teaches wherein the loss function is a loss function obtained by comparing losses due to differences in distance between the extracted teacher sparse activation map and the extracted student sparse activation map. (“At step S130, distances between corresponding feature vectors in the first feature map and the second feature map are calculated, corresponding elements in the segmentation labeling factor matrix are used to correct the distances, so as to obtain corrected distances between corresponding feature vectors in the first feature map and the second feature map, and a sum of the corrected distances between all the feature vectors in the first feature map and the second feature map is calculated as the first loss function.” [¶0066])
Regarding claim 8, Han/Heo teaches The method of claim 7, where Heo teaches
PNG
media_image3.png
238
630
media_image3.png
Greyscale
(“
PNG
media_image4.png
104
330
media_image4.png
Greyscale
” [pg. 2, eq(1); note: σ(T(I) corresponds to teacher network activation map and σ(S(I)) corresponds to student network activation map. Note: Eq(7) further shows the summation of the teacher/student activations.])
Same motivation to combine the teachings of Han/Heo as claim 1.
Regarding claims 9, 10, 12, and 15-16, they are substantially similar to claim 1-2, 4, 7-8 respectively, and are rejected in the same manner, the same art, and reasoning applying.
Claims 3, 5, 11 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Han in view of Heo and further in view of Li et al. ("Knowledge Distillation with Attention for Deep Transfer Learning of Convolutional Networks", hereinafter "Li").
Regarding claim 3, Han/Heo teaches The method of claim 2, however fails to explicitly teach wherein the extracting of the teacher sparse activation maps further comprises:
normalizing the activation map extracted from the learning model of the teacher network to a preset range; and re-filtering the normalized activation map through the activation function.
Li teaches normalizing the activation map extracted from the learning model of the teacher network to a preset range; (“We select some convolution filters on which the source model (the initialization before fine-tuning) has low activation. For the convenience of analyzing the effect of regularization methods, each element ai of the original activation map is normalized with… where the min and max terms in the formula represent for the minimum and maximum value of the whole activation map, respectively” [pg. 13, ¶3]) and
re-filtering the normalized activation map through the activation function. (“To further understand the effect of attention and the implication of “unactivated channel re usage”, we “attribute” the visual attention to the original image to identify the set of pixels having high contributions in the activated feature maps.” [pg. 13, ¶3]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Han’s/Heo’s teachings to implement normalization and re-filtering as taught by Li. One would have been motivated to make this modification in order to understand the effect of attention and identify high activation at important regions in the activated feature maps. [pg. 13, ¶3, Li]
Regarding claim 5, Han/Heo teaches The method of claim 4, however fails to explicitly teach wherein the extracting of the student sparse activation maps further comprises:
normalizing the activation map extracted from the learning model of the student network to a preset range; and re-filtering the normalized activation map through the activation function.
Li teaches normalizing the activation map extracted from the learning model of the teacher network to a preset range; (“We select some convolution filters on which the source model (the initialization before fine-tuning) has low activation. For the convenience of analyzing the effect of regularization methods, each element ai of the original activation map is normalized with… where the min and max terms in the formula represent for the minimum and maximum value of the whole activation map, respectively” [pg. 13, ¶3]) and
re-filtering the normalized activation map through the activation function. (“To further understand the effect of attention and the implication of “unactivated channel re usage”, we “attribute” the visual attention to the original image to identify the set of pixels having high contributions in the activated feature maps.” [pg. 13, ¶3]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Han’s/Heo’s teachings to implement normalization and re-filtering as taught by Li. One would have been motivated to make this modification in order to understand the effect of attention and identify high activation at important regions in the activated feature maps. [pg. 13, ¶3, Li]
Regarding claims 11 and 13, they are substantially similar to claims 3 and 5 respectively, and are rejected in the same manner, the same art, and reasoning applying.
Allowable Subject Matter
Claims 6 and 14 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Regarding claims 6 and 14:
wherein the activation function is a rectified linear unit (ReLU) function, and the filter function may be a function obtained by moving the coordinates of the ReLU function and adjusting a passband.
No prior art was uncovered which fairly discloses obtaining the filter function by moving coordinates of the ReLU function and adjusting a passband.
The closest prior art cited is Han et al. (“US 20260004138 A1”) which discloses a training method for a defect detection model using knowledge distillation. However, the reference does not go into details of moving coordinates of the ReLU function and adjusting a passband to obtain a filter function.
Heo et al. (“Knowledge Transfer via Distillation of Activation Boundaries
Formed by Hidden Neurons”) discloses a method for knowledge transfer via distillation of activation boundaries of neurons. However, the reference does not disclose any details regarding moving coordinates of the ReLU function and adjusting a passband to obtain a filter function.
Li et al. (“Knowledge Distillation with Attention for Deep Transfer
Learning of Convolutional Networks”) discloses a transfer learning method using feature maps with attention. However, the reference does not explicitly disclose any details regarding moving coordinates of the ReLU function and adjusting a passband to obtain a filter function.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MICHAEL H HOANG whose telephone number is (571)272-8491. The examiner can normally be reached Mon-Fri 8:30AM-4:30PM.
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, Kakali Chaki can be reached at (571) 272-3719. 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.
/MICHAEL H HOANG/ PRIMARY EXAMINER, Art Unit 2122
Prosecution Insights