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 .
Filed Amendments/Remarks after Restriction/Election Requirement of 01/12/2024 have been entered.
Applicant’s elected claims 1-9, and 20.
Claims 10-19 have been cancelled.
New claims 21-29 have been added.
Currently claims 1-9, and 20-29 are pending.
Please refer to the action below.
Examiner Notes
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. However, the claimed subject matter, not the specification, is the measure of the invention.
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 1-9, and 20-29 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim(s) 1, recite(s) mental processes and software processes directed to a system, a semiconductor specimen, and a processing circuitry. Claim 1 (being the representative claim) further recites similar limitations to those discussed with regards to respective independent claims 20, and 21, and therefore discussion is omitted for brevity.
Independent claim 1 includes limitations that recite an abstract idea. Claim 1 recites: A computerized system for runtime defect examination on a semiconductor specimen, the system comprising a processing circuitry configured to: obtain an input image indicative of difference between an inspection image of the specimen and a corresponding reference image; and process the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image, wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image based on Fourier transform, and wherein the output image is usable for further defect examination.
The claim recites the steps of obtain a difference in put image between an inspection image of the specimen and a corresponding reference image, and similarly computing a defect map indicative of distribution of defect of interest (DOI) candidates in the input image include steps which further falls within the mathematical concepts grouping of abstract ideas where the difference image and generated defect map are generated by performing mathematical calculations. The claims also recite the steps of process the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image, wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image based on Fourier transform, and wherein the output image is usable for further defect examination which steps may be further performed practically in the human mind as mental processes of observing the input images in the trained machine learning system to identify location and/or presence of image defect and to output the defect map indicative of observed distribution of defect of interest (DOI) candidates in the input image.
Thus, the claim recites at least a Fourier transform calculation mathematical as well as a mathematical computing difference calculation, both of which fall within the mathematical concepts grouping of abstract ideas. As explained in the MPEP, when a claim recites multiple abstract ideas that fall in the same or different groupings, examiners should consider the limitations together as a single abstract idea, rather than as a plurality of separate abstract ideas to be analyzed individually. See MPEP 2106.04, subsection II.B. As the steps (b) and (c) fall within the same grouping of abstract ideas (i.e., mathematical concepts), these limitations are considered together as a single abstract idea for further analysis. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application because the claims merely recite mental steps that can be performed by a person and/or software steps that can be performed by component or units of a software. That is, other than reciting “process the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image, wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image based on Fourier transform, and wherein the output image is usable for further defect examination” nothing in the claim element precludes the steps from practically being performed in the mind and/or purely by software. The additional elements of “process the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image, wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image based on Fourier transform, and wherein the output image is usable for further defect examination” does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. Hence, claim 1 is not subject matter eligible.
The dependent claims 2-9, and 22-29 do not recite any further limitations that cause the claim(s) to be subject matter eligible. Rather, the limitations of dependent claims are directed toward additional aspects of the judicial exception and/or well-understood, routine and conventional additional elements that do not integrate the judicial exception into a practical application. Based on broadest reasonable interpretation of the claims, all of the steps recited in the independent claims 1, and 20-21 corresponding to dependent claims 2-9, and 22-29 further correspond to concepts performed by at least software components which may be further performed in the human mind. Additionally, a person mentally in the human mind and/or software can perform the obtaining of a difference image, inputting the said image in a machine learning system to produce and output image defect map indicative of distribution of defect of interest (DOI) candidates in the input image and to continue using said output image for further defect examination. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
Concepts performed in the human mind have been identified in the 2019 PEG as an exemplar in the “Mental Process” grouping of abstract ideas. For the reasons above, the claims do not amount to significantly more than an abstract idea. Even when considered in combination, these additional elements represent mere instructions to apply an exception and insignificant extra-solution activity, which do not provide an inventive concept and therefore, the claims are not patent-eligible.
Furthermore, these additional generic hardware elements perform no more than their basic computer function. Generic computer‐implementation of a method is not a meaningful limitation that alone can amount to significantly more than an abstract idea. Moreover, when viewed as a whole with such additional element considered as an ordered combination, claims modified by adding generic hardware elements are nothing more than a purely conventional computerized implementation of an idea in the general field of computer processing and do not provide significantly more than an abstract idea.
Consequently, the identified additional generic hardware elements taken into consideration individually and in combination fail to amount to significantly more than the abstract idea 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 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 difference the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 7-9, 20-21, and 27-29 is/are rejected under 35 U.S.C. 103 as obvious over Cohen et al. (US 2020/0226744, A1), in view of Huang et al. (US 2014/0270475, A1).
Regarding claim 1, Cohen teaches a computerized system for runtime defect examination on a semiconductor specimen (at least para. 0009 and 0086-0088 of the disclosure teaches a computerized system for runtime defect examination on a semiconductor specimen),
the system comprising a processing circuitry configured to:
obtain an input image indicative of difference between an inspection image of the specimen and a corresponding reference image (obtaining further in at least para. 0082 a difference input image indicative of difference between an inspection image of the specimen and a corresponding reference image); and
process the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image (processing further in at least 0088 using a trained machine learning (ML) system the input image of further para. 0082-0084 to generate an output image representative of a noise/defect map of para. 0047-0048 and 0082-0084 indicative of distribution of defect of interest (DOI) candidates in the input image),
wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image
and wherein the output image is usable for further defect examination (at least Fig. 2, S212 further teaches the generated noise map image is usable for further defect examination).
However, Cohen is silent regarding the above lined-out items such as wherein said plurality of ML models operatively connected therebetween and previously trained together to perform said defect detection on the input image specifically based on Fourier transform.
Huang teaches at least in para. 0002-0009 a system for detecting defects on a semiconductor wafer, the system further in para. 0002-0009 teaches the generating of at least one or more output defect maps images based on at least obtained or generated difference image data of further para. 0022, the system further in at least para. 0047 and 0063 further adapted for generating output image data based on a Fourier filtering method for further performing defect detection on input image and/or outputting the output defect map images as said Fourier Filtering is known in the art for using Fourier transform to further process said image data or output image. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang to include wherein said plurality of ML models operatively connected therebetween and previously trained together to perform said defect detection on the input image specifically based on Fourier transform, as discussed above, as Cohen in view of Huang are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Huang ’s image defect detection of the input image based on Fourier transform and that of the output defect maps image further complements the machine learning generated defect map output image of Cohen in the sense that said image defect detection based on Fourier Filtering of Huang when combined with Cohen’s machine generated defect maps facilitates further defect and residual detection corresponding to generated defect maps using at least Fourier filtering known in the art to use Fourier transform corresponding to the input image which as cited by Huang eliminate or suppress residual noises/nuisance using said Fourier filtering which as described herein by Huang “may be used to enhance the detectability of DOI for wafer inspection systems and provide increased defect detection sensitivity”, according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Regarding claim 7 (according to claim 1), Cohen further teaches wherein the processing circuitry is further configured to: obtain a second input image indicative of difference between the inspection image and a second reference image
(at least para. 0081-0082 further teaches the system as implied iteratively obtain one or more second input image indicative of difference between the inspection image and a second reference image);
process the second input image using the trained ML system to obtain a second output image (the system as implied further in at least para. 0081-0082 further \ iteratively process a second or more difference input image using the trained ML system of further para. 0088 to obtain a second output image); and
determine presence of DOI candidates based on the output image and the second output image (and determining further in at least para. 0047-0048 presence of DOI candidates based on the output image and the second output image of further Fig. 2).
Regarding claim 8 (according to claim 1), Cohen further teaches wherein the ML system comprises a classifier (para. 0004 of the disclosure further cites “the term “examination” used in this specification should be expansively construed to cover any kind of detection and/or classification of defects in an object” further indicating the system impliedly comprises at least a defect classifier);
and the processing circuitry is further configured to use the classifier to process the output image to provide a classification score indicative of level of confidence of DOI presence in the output image, and determine DOI presence based on the output image and the classification score (at least the Abstract teaches a calculated score for a detected region of the given noise map further indicative of the process output image to provide as implied further in at least para. 0004 an examination/classification score indicative in the art of level of confidence of DOI presence shown in at least para. 0047-0048 in the output image, and configured further to determine DOI presence based on the output image and the classification score).
Regarding claim 9 (according to claim 1), Cohen further teaches wherein the input image is a difference image obtained by comparing the inspection image with the reference image (para. 0081-0084).
However, Cohen is silent regarding wherein the output image has suppressed residual noises with respect to the input image, which, when being used for further defect examination, improves detection sensitivity.
Huang teaches further in at least para. 0043-0044 suppressing residual noises with respect to the input image to produce inherently a suppressed noise residual output image with respect to the input image, which, when being used for further defect examination, improves detection sensitivity of further para. 0007. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang to include wherein said output image has suppressed residual noises with respect to the input image, which, when being used for further defect examination, improves detection sensitivity, as discussed above, as Cohen in view of Huang are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Huang ’s image defect and/or residuals suppression further complements the machine learning generated defect map output image of Cohen in the sense that said image defect and/or residuals suppression of Huang when combined with Cohen’s machine generated defect maps facilitates further defect and residual detection corresponding to generated defect maps using at least Fourier filtering known in the art to use Fourier transform corresponding to the input image which as cited by Huang eliminate or suppress residual noises/nuisance using said Fourier filtering which as described herein by Huang “may be used to enhance the detectability of DOI for wafer inspection systems and provide increased defect detection sensitivity”, according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Regarding claim 20, Cohen teaches in at least para. 0025 a non-transitory computer readable storage medium tangibly embodying a program of instructions that, when executed by a computer, cause the computer to perform a method of runtime defect examination on a semiconductor specimen,
the method comprising:
obtaining an input image indicative of difference between an inspection image of the specimen and a corresponding reference image (obtaining further in at least para. 0082 a difference input image indicative of difference between an inspection image of the specimen and a corresponding reference image); and
processing the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image (processing further in at least 0088 using a trained machine learning (ML) system the input image of further para. 0082-0084 to generate an output image representative of a noise/defect map of para. 0047-0048 and 0082-0084 indicative of distribution of defect of interest (DOI) candidates (para. 0047-0048) in the input image),
wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image
and wherein the output image is usable for further defect examination (at least Fig. 2, S212 further teaches the generated noise map image is usable for further defect examination).
However, Cohen is silent regarding the above lined-out items such as wherein said plurality of ML models operatively connected therebetween and previously trained together to perform said defect detection on the input image specifically based on Fourier transform.
Huang teaches at least in para. 0002-0009 a system for detecting defects on a semiconductor wafer, the system further in para. 0002-0009 teaches the generating of at least one or more output defect maps images based on at least obtained or generated difference image data of further para. 0022, the system further in at least para. 0047 and 0063 further adapted for generating output image data based on a Fourier filtering method for further performing defect detection on input image and/or outputting the output defect map images as said Fourier Filtering is known in the art for using Fourier transform to further process said image data or output image. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang to include wherein said plurality of ML models operatively connected therebetween and previously trained together to perform said defect detection on the input image specifically based on Fourier transform, as discussed above, as Cohen in view of Huang are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Huang ’s image defect detection of the input image based on Fourier transform and that of the output defect maps image further complements the machine learning generated defect map output image of Cohen in the sense that said image defect detection based on Fourier Filtering of Huang when combined with Cohen’s machine generated defect maps facilitates further defect and residual detection corresponding to generated defect maps using at least Fourier filtering known in the art to use Fourier transform corresponding to the input image which as cited by Huang eliminate or suppress residual noises/nuisance using said Fourier filtering which as described herein by Huang “may be used to enhance the detectability of DOI for wafer inspection systems and provide increased defect detection sensitivity”, according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Regarding claim 21, Cohen teaches in at least para. 0023 a method for runtime defect examination on a semiconductor specimen,
the method comprising:
obtaining an input image indicative of difference between an inspection image of the specimen and a corresponding reference image (obtaining further in at least para. 0082 a difference input image indicative of difference between an inspection image of the specimen and a corresponding reference image); and
processing the input image using a trained machine learning (ML) system, to generate an output image representative of a defect map indicative of distribution of defect of interest (DOI) candidates in the input image (processing further in at least 0088 using a trained machine learning (ML) system the input image of further para. 0082-0084 to generate an output image representative of a noise/defect map of para. 0047-0048 and 0082-0084 indicative of distribution of defect of interest (DOI) candidates (para. 0047-0048) in the input image),
wherein the ML system comprises a plurality of ML models operatively connected therebetween and previously trained together to perform defect detection on the input image
and wherein the output image is usable for further defect examination (at least Fig. 2, S212 further teaches the generated noise map image is usable for further defect examination).
However, Cohen is silent regarding the above lined-out items such as wherein said plurality of ML models operatively connected therebetween and previously trained together to perform said defect detection on the input image specifically based on Fourier transform.
Huang teaches at least in para. 0002-0009 a system for detecting defects on a semiconductor wafer, the system further in para. 0002-0009 teaches the generating of at least one or more output defect maps images based on at least obtained or generated difference image data of further para. 0022, the system further in at least para. 0047 and 0063 further adapted for generating output image data based on a Fourier filtering method for further performing defect detection on input image and/or outputting the output defect map images as said Fourier Filtering is known in the art for using Fourier transform to further process said image data or output image. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang to include wherein said plurality of ML models operatively connected therebetween and previously trained together to perform said defect detection on the input image specifically based on Fourier transform, as discussed above, as Cohen in view of Huang are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Huang ’s image defect detection of the input image based on Fourier transform and that of the output defect maps image further complements the machine learning generated defect map output image of Cohen in the sense that said image defect detection based on Fourier Filtering of Huang when combined with Cohen’s machine generated defect maps facilitates further defect and residual detection corresponding to generated defect maps using at least Fourier filtering known in the art to use Fourier transform corresponding to the input image which as cited by Huang eliminate or suppress residual noises/nuisance using said Fourier filtering which as described herein by Huang “may be used to enhance the detectability of DOI for wafer inspection systems and provide increased defect detection sensitivity”, according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Regarding claim 27 (according to claim 21), Cohen further teaches wherein further comprising: obtaining a second input image indicative of difference between the inspection image and a second reference image
(at least para. 0081-0082 further teaches the system as implied iteratively obtain one or more second input image indicative of difference between the inspection image and a second reference image);
processing the second input image using the trained ML system to obtain a second output image (the system as implied further in at least para. 0081-0082 further iteratively process a second or more difference input image using the trained ML system of further para. 0088 to obtain a second output image); and
determining presence of DOI candidates based on the output image and the second output image (and determining further in at least para. 0047-0048 presence of DOI candidates based on the output image and the second output image of further Fig. 2).
Regarding claim 28 (according to claim 21), Cohen further teaches wherein the ML system comprises a classifier (para. 0004 of the disclosure further cites “the term “examination” used in this specification should be expansively construed to cover any kind of detection and/or classification of defects in an object” further indicating the system impliedly comprises at least a defect classifier);
and the method further comprises using the classifier to process the output image to provide a classification score indicative of level of confidence of DOI presence in the output image, and determining DOI presence based on the output image and the classification score (at least the Abstract teaches a calculated score for a detected region of the given noise map further indicative of the process output image to provide as implied further in at least para. 0004 an examination/classification score indicative in the art of level of confidence of DOI presence shown in at least para. 0047-0048 in the output image, and configured further to determine DOI presence based on the output image and the classification score).
Regarding claim 29 (according to claim 21), Cohen further teaches wherein the input image is a difference image obtained by comparing the inspection image with the reference image (para. 0081-0084).
However, Cohen is silent regarding wherein the output image has suppressed residual noises with respect to the input image, which, when being used for further defect examination, improves detection sensitivity.
Huang teaches further in at least para. 0043-0044 suppressing residual noises with respect to the input image to produce inherently a suppressed noise residual output image with respect to the input image, which, when being used for further defect examination, improves detection sensitivity of further para. 0007. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang to include wherein said output image has suppressed residual noises with respect to the input image, which, when being used for further defect examination, improves detection sensitivity, as discussed above, as Cohen in view of Huang are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Huang ’s image defect and/or residuals suppression further complements the machine learning generated defect map output image of Cohen in the sense that said image defect and/or residuals suppression of Huang when combined with Cohen’s machine generated defect maps facilitates further defect and residual detection corresponding to generated defect maps using at least Fourier filtering known in the art to use Fourier transform corresponding to the input image which as cited by Huang eliminate or suppress residual noises/nuisance using said Fourier filtering which as described herein by Huang “may be used to enhance the detectability of DOI for wafer inspection systems and provide increased defect detection sensitivity”, according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Claims 5, and 25 is/are rejected under 35 U.S.C. 103 as obvious over Cohen in view of Huang, and further in view of Bergmann et al. (EP 4145401, A1).
Regarding claim 5 (according to claim 1), Cohen further teaches in at least para. 0088 wherein the ML system comprises a first ML model operatively connected to a second model.
However, Cohen in view of Huang are silent regarding wherein the first ML model is configured to generate one or more feature maps representative of the input image, and the second model is configured to process the one or more feature maps and reconstruct the output image.
Bergmann teaches in at least para 0115 the used machine learning system to detect image anomalies and to output a subsequent feature maps image based on obtained input images, Bergmann further teaches an auto-encoder model as at least one or more model of the disclosure trained to process the one or more feature maps and reconstruct the output feature map image through as cited “a lower-dimensional intermediate representation to match a target representation”. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang, and further in view of Bergmann to include wherein said first ML model is configured to generate one or more feature maps representative of the input image, and the second model is configured to process the one or more feature maps and reconstruct the output image, as discussed above, as Cohen in view of Huang, and further in view of Bergmann are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Bergmann’ s image defect detection and image reconstruction corresponding to output feature maps further complements the machine learning generated defect map output image of Cohen in view of Huang in the sense that said image defect detection and image reconstruction corresponding to output feature maps of Bergmann when combined with Cohen in view of Huang machine generated defect maps optimizes target image reconstruction corresponding to the input image according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Regarding claim 25 (according to claim 21), Cohen further teaches in at least para. 0088 wherein the ML system comprises a first ML model operatively connected to a second model.
However, Cohen in view of Huang are silent regarding wherein the first ML model is configured to generate one or more feature maps representative of the input image, and the second model is configured to process the one or more feature maps and reconstruct the output image.
Bergmann teaches in at least para 0115 the used machine learning system to detect image anomalies and to output a subsequent feature maps image based on obtained input images, Bergmann further teaches an auto-encoder model as at least one or more model of the disclosure trained to process the one or more feature maps and reconstruct the output feature map image through as cited “a lower-dimensional intermediate representation to match a target representation”. It would have been obvious to one of ordinary skill in the art at the time the invention was made to combine the teachings of Cohen in view of Huang, and further in view of Bergmann to include wherein said first ML model is configured to generate one or more feature maps representative of the input image, and the second model is configured to process the one or more feature maps and reconstruct the output image, as discussed above, as Cohen in view of Huang, and further in view of Bergmann are in the same field of performing runtime inspection of semiconductor specimen or wafers so as to detect defects on an input image and to generate an output image based on the detection result, Bergmann’ s image defect detection and image reconstruction corresponding to output feature maps further complements the machine learning generated defect map output image of Cohen in view of Huang in the sense that said image defect detection and image reconstruction corresponding to output feature maps of Bergmann when combined with Cohen in view of Huang machine generated defect maps optimizes target image reconstruction corresponding to the input image according to further known methods to yield predictable results since known work in one field of endeavor may prompt variations of it for use in either the same field or a different one based on design incentives or other market forces if the variations are predictable to one of ordinary skill in the art as said combination is thus the adaptation of an old idea or invention using newer technology that is either commonly available and understood in the art thereby a variation on already known art (See MPEP 2143, KSR Exemplary Rationale F).
Claims Standings
Dependent claims 2-4, 6, 22-24, 26 are objected over the prior arts of record, 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, and if all outstanding rejections are overcome.
The prior arts do not appear to teach: claim 2. (Original) The computerized system according to claim 1, wherein the ML system comprises a first ML model operatively connected to a second ML model and a third ML model in parallel, wherein the first ML model is configured to generate one or more feature maps representative of the input image, and the second and third ML models are configured to process the one or more feature maps separately and generate, respectively, a real part and an imaginary part of a Fourier image corresponding to the input image.
3. (Original) The computerized system according to claim 2, wherein the ML system further comprises an analytical model configured to perform inverse Fourier transform (IFT) on the Fourier image to reconstruct the output image.
4. (Original) The computerized system according to claim 2, wherein the first ML model is an auto-encoder, and the second ML model and the third ML model are Convolutional Neural Networks (CNNs).
6. (Original) The computerized system according to claim 1, wherein the ML system is trained to detect presence of a DOI in a training image based on a frequency response of a Fourier image corresponding to the training image and a ground truth frequency response of the DOI, and identify a location of the DOI in the training image based on the training image and a reconstructed image of the training image.
22. (New) The method of claim 21, wherein the ML system comprises a first ML model operatively connected to a second ML model and a third ML model in parallel, wherein the first ML model is configured to generate one or more feature maps representative of the input image, and the second and third ML models are configured to process the one or more feature maps separately and generate, respectively, a real part and an imaginary part of a Fourier image corresponding to the input image.
23. (New) The method of claim 22, wherein the ML system further comprises an analytical model configured to perform inverse Fourier transform (IFT) on the Fourier image to reconstruct the output image.
24. (New) The method of claim 22, wherein the first ML model is an auto-encoder, and the second ML model and the third ML model are Convolutional Neural Networks (CNNs).
26. (New) The method of claim 21, wherein the ML system is trained to detect presence of a DOI in a training image based on a frequency response of a Fourier image corresponding to the training image and a ground truth frequency response of the DOI, and identify a location of the DOI in the training image based on the training image and a reconstructed image of the training image.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARCELLUS AUGUSTIN whose telephone number is (571)270-3384. The examiner can normally be reached 9 AM- 5 PM.
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, BENNY TIEU can be reached at 571-272-7490. 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.
/MARCELLUS J AUGUSTIN/Primary Examiner, Art Unit 2682 01/22/2026