DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . Claims 1-20 are pending.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more.
Step 1 (The Statutory Categories): Is the claim to a process, machine, manufacture, or composition of matter? MPEP 2106.03.
Per Step 1, claims 1 and 16 are is to a non-transitory computer-readable medium (i.e., a manufacture), claim 19 to a system (i.e., a machine). Thus, the claims are directed to statutory categories of invention. However, the claims are rejected under 35 U.S.C. 101 because they are directed to an abstract idea, a judicial exception, without reciting additional elements that integrate the judicial exception into a practical application.
The analysis proceeds to Step 2A Prong One.
(Examiner notes that claim 19, while dependent on claim 16, claims a “system” and is therefore treated as an independent claim and separate statutory category in the analysis below. Regardless of how it’s interpreted, whether independent or dependent, it’s still ineligible.)
Step 2A Prong One: Does the claim recite an abstract idea, law of nature, or natural phenomenon? MPEP 2106.04.
The abstract idea of claim 1 is:
access a first SEM data set and a second SEM data set , the first SEM data set and the second SEM data set being associated with a patterned substrate; and
train, using the first SEM data set and the second SEM data set as training data, a model such that the trained model is configured to convert a metrology data set acquired by the second SEM system to a converted data set having characteristics comparable to metrology data being acquired by the first SEM system.
The abstract idea of claim 16 is:
convert the captured metrology data via a trained model into converted metrology data, the converted metrology data having characteristics as if captured by a second metrology system.
The abstract idea of claim 19 is (brackets indicate abstract steps inherited from claim 16):
detect a substrate;
[convert the captured metrology data via a trained model into converted metrology data, the converted metrology data having characteristics as if captured by a second metrology system.]
The abstract idea steps italicized above are those which could be performed mentally, including with pen and paper. The steps describe, at a high level: accessing SEM data sets and training using the SEM data sets to convert a metrology data set; or detecting a substrate and/or converting captured metrology data via a trained model into converted metrology data. (Examiner notes that the “training” and “trained model” may be accomplished via simple linear regression using pen and paper.) If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind, including observations, evaluations, judgements, and/or opinions, then it falls within the Mental Processes – Concepts Performed in the Human Mind grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Additionally and alternatively, the abstract idea steps italicized above describe the rules or instructions pertaining to converting a metrology data set, which constitutes a process that, under its broadest reasonable interpretation, covers managing personal behavior relationships, interactions between people. If a claim limitation, under its broadest reasonable interpretation, covers managing personal behavior relationships, interactions between people, including social activities, teaching, and/or following rules or instructions, then it falls within the Certain Methods of Organizing Human Activity – Managing Personal Behavior Relationships, Interactions Between People grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Additionally and alternatively, the abstract idea steps italicized above describe accessing SEM data sets and training using the SEM data sets to convert a metrology data set; or detecting a substrate and/or converting captured metrology data via a trained model into converted metrology data, which constitutes a process that, under its broadest reasonable interpretation, covers mathematical concepts. (Examiner notes that the “training” and “trained model” may be accomplished via simple linear regression using pen and paper.) If a claim limitation, under its broadest reasonable interpretation, covers mathematical concepts, including mathematical relationships, mathematical formulas or equations, mathematical calculations, then it falls within the Mathematical Concepts grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong Two: Does the claim recite additional elements that integrate the judicial exception into a practical application? MPEP 2106.04.
This judicial exception is not integrated into a practical application because the additional elements are merely instructions to apply the abstract idea to a computer, as described in MPEP 2106.05(f).
Claim 1 recites the following additional elements: a non-transitory computer-readable medium having instructions recorded therein or thereon, the instructions, when executed by a computer system, configured to cause the computer system; acquired by a first scanning electron metrology (SEM) system; acquired by a second SEM system; machine learning (ML); ML.
Claim 16 recites the following additional elements: a non-transitory computer-readable medium having instructions recorded therein or thereon, the instructions, when executed by a computer system, configured to cause the computer system; capture metrology data, obtained by a first metrology system, of a patterned substrate; ML.
Claim 19 recites the following additional elements: a detector configured to; a computer system including one or more processors. (Additional elements inherited from claim 16 include: a non-transitory computer-readable medium having instructions recorded therein or thereon, the instructions, when executed by a computer system, configured to cause the computer system; capture metrology data, obtained by a first metrology system, of a patterned substrate; ML.)
These elements – with the exception of “acquired by a first scanning electron metrology (SEM) system,” “acquired by a second SEM system,” and “capture metrology data, obtained by a first metrology system, of a patterned substrate” – are merely instructions to apply the abstract idea to a computer, per MPEP 2106.05(f). Applicant has only described generic computing elements in their specification, as seen in [0093]-[00101] of applicant’s specification as filed, for example.
The remaining additional elements – “acquired by a first scanning electron metrology (SEM) system,” “acquired by a second SEM system,” and “capture metrology data, obtained by a first metrology system, of a patterned substrate” – merely elaborate on the field of use, i.e., metrology systems. Simply linking to a technology or field of use doesn’t integrate into practical application. See MPEP 2106.05(h).
Further, the combination of these elements is nothing more than a generic computing system applied to the tasks of the abstract idea, one tied to a field of use. Because the additional elements are merely instructions to apply the abstract idea to a generic computing system and/or generally link to a field of use, they do not integrate the abstract idea into a practical application, when viewed alone or in combination. See MPEP 2106.05(f), (h).
Therefore, per Step 2A Prong Two, the additional elements, alone and in combination, do not integrate the judicial exception into a practical application. The claim is directed to an abstract idea.
Step 2B (The Inventive Concept): Does the claim recite additional elements that amount to significantly more than the judicial exception? MPEP 2106.05.
Step 2B involves evaluating the additional elements to determine whether they amount to significantly more than the judicial exception itself.
The examination process involves carrying over identification of the additional element(s) in the claim from Step 2A Prong Two and carrying over conclusions from Step 2A Prong Two pertaining to MPEP 2106.05(f), (h).
The additional elements and their analysis are therefore carried over: applicant has merely recited elements that facilitate the tasks of the abstract idea, as described in MPEP 2106.05(f), and/or generally linked to a field of use, as described in MPEP 2106.05(h).
Further, the combination of these elements is nothing more than a generic computing system applied to the tasks of the abstract idea, one tied to a field of use. Because the additional elements are merely instructions to apply the abstract idea to a generic computing system and/or generally link to a field of use, they do not amount to significantly more, when viewed alone or in combination. See MPEP 2106.05(f), (h).
Therefore, per Step 2B, the additional elements, alone and in combination, are not significantly more. The claims are not patent eligible.
The analysis takes into consideration all dependent claims as well:
Dependent claims 2-15, 17-18, and 20 recite additional abstract steps and/or information that further narrow the abstract idea(s) above. This narrowing of the abstract idea does not integrate it into practical application and/or add significantly more.
Some of the dependent claims recite further additional elements, beyond those highlighted above:
Claim 15: wherein the ML model is trained using a generative adversarial network (GAN) architecture, the ML model comprising a generator model and a discriminator model.
Claim 17: machine learning (ML) model.
Claim 20: wherein the metrology system is a scanning electron microscope.
Similar to above, these are generic computing elements used to facilitate the tasks of the narrowed abstract idea(s). Whether viewed alone or in combination, these further additional elements do not integrate the narrowed abstract idea(s) into practical application and/or add significantly more.
Accordingly, claims 1-20 are rejected under 35 USC § 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-5, 9-12, and 16-20 are rejected under 35 U.S.C. 102(a)(1), (a)(2) as being anticipated by Mossavat (US 20200110341).
Claim 1
Mossavat discloses:
A non-transitory computer-readable medium having instructions recorded therein or thereon, the instructions, when executed by a computer system {See [0083].}, configured to cause the computer system to at least:
access a first SEM data set acquired by a first scanning electron metrology (SEM) system and a second SEM data set acquired by a second SEM system, the first SEM data set and the second SEM data set being associated with a patterned substrate {first, second SEM data sets from first, second SEM systems accessed, where first, second SEM data sets associated with patterned substrate(s) described in [0047]: In an embodiment, the training data 32 is obtained by measuring the same substrate W or set of substrates W with all of the multiple metrology apparatuses 30.sub.1-30.sub.N to be matched. In the example shown, the training data 32 is obtained by performing metrology measurements on two different sets 34.sub.1 and 34.sub.2 of substrates. Metrology apparatuses 30.sub.1-30.sub.3 contribute to the training data 32 by measuring substrates W in substrate set 34.sub.1 and metrology apparatuses 30.sub.4-30.sub.N contribute to the training data 32 by measuring substrates W in substrate set 34.sub.2. patterned substrate also described in [0025].}; and
train, using the first SEM data set and the second SEM data set as training data, a machine learning (ML) model such that the trained ML model is configured to convert a metrology data set acquired by the second SEM system to a converted data set having characteristics comparable to metrology data being acquired by the first SEM system {training using first, second SEM data sets and converting second SEM data set having comparable characteristics to that acquired by first SEM system described in [0079]: In any of the embodiments discussed above, an initial training of the first machine learning process 41 and the second machine learning process 42 may be performed using detected representations obtained from a set of metrology apparatuses 30.sub.1-30.sub.N over a set of representative applications. In an alternative approach, a selected one metrology apparatus may be used as a reference for training other metrology apparatuses to achieve matching to the reference metrology apparatus. This training may be performed using a transfer-learning technique. In an embodiment, as depicted in FIG. 9, a virtual metrology apparatus 30.sub.V is used as the reference. Training target profiles 50 are provided to the virtual metrology apparatus 30.sub.V. Simulation of measurement of the training target profiles by the virtual metrology apparatus 30.sub.V provides datasets DS to be input to the encoder F.sub.1 to provide encoded representations CD. This approach may improve the accuracy of the metrology apparatus 30. It is desirable that the training target profiles 50 used for simulation, at least for parameters of interest, cover the distribution of the actual substrates W to be measured, since all differences will be detected by the classifier CL and used to train the encoder/decoder combination to remove these differences from the encoded representation.}.
Claim 2
Mossavat further discloses: wherein the first and second SEM data sets comprise SEM images of the patterned substrate {SEM images of the patterned substrate described in [0041], [0042].}, and wherein the instructions configured to cause the computer system to train the ML model {training described in [0051]-[0055]} are further configured to cause the computer system to: compare the first set of images acquired by the first SEM system and second set of images acquired by the second SEM system {comparing images described in [0066]: Two competing training mechanisms (the first machine learning process 41 and the second machine learning process 42) are used to optimize the network 44. In this example, one comparator 48.sub.1-48.sub.N is provided per encoder F.sub.1. The comparators 48.sub.1-48.sub.N compare the datasets DS input to the encoders F.sub.1 with the datasets MS output from the decoder F.sub.2 and provides feedback to adjust parameters defining the encoders F.sub.1 to optimize the cost function 43 (see broken line data paths) and thereby attempt to maximize the preservation of information in the datasets MS relative to the datasets DS. The classifier CL also receives the datasets MS from the decoder F.sub.2 and will be trained to optimize the cost function 43 and thereby attempt to maximize the probability of the classifier CL classifying each dataset MS to the correct corresponding metrology apparatus 30.sub.1-30.sub.N. Data flow for training of the classifier is indicated by thick solid lines.}; and adjust one or more parameters of the ML model based on the comparison to influence a cost function used to train the ML model to improve matching between the first set of images and ML-generated images using the second set of images as input to the ML model {See previous citation to [0066]}.
Claim 3
Mossavat further discloses: wherein the first and second SEM data sets comprise: contours of features on the patterned substrate; and/or a physical characteristic associated with patterns on the patterned substrate {physical characteristics described in [0063]: Thus, the encoded representation z may comprise reconstructed geometrical dimensions of the geometrical model, e.g., critical dimension, side wall angle, overlay, etc. In such an embodiment, the training of the first machine learning process 41 may comprise adjusting parameters (e.g., material parameters, nominal stack dimensions, fix/float, etc.) defining the geometrical model (i.e. the geometrical model is parametrized by θ.sub.1) and/or adjusting one or more parameters defining the metrology recipe (i.e. the metrology recipe is parametrized by θ.sub.1). Examiner notes that broadest reasonable interpretation requires consideration of only one alternative.}.
Claim 4
Mossavat further discloses: wherein the first and second SEM data sets comprise a physical characteristic associated with patterns on the patterned substrate and wherein the physical characteristic comprises critical dimension (CD) of the patterns on the patterned substrate {See previous citation to [0063] for physical characteristic. Also see [0029] for critical dimension (CD): A metrology apparatus, which may also be referred to as an inspection apparatus, is used to measure properties of targets on substrates W, such as overlay error (OV), critical dimension (CD), or more complex shape parameters.}.
Claim 5
Mossavat further discloses: wherein the instructions configured to cause the computer system to train the ML model are further configured to cause the computer system to: compare first CD values of the first SEM data set and second CD values of the second SEM data set {comparing CD values described in [0063]: In an embodiment, the encoder F.sub.1 derives one or more target parameters of a geometrical model of the structure on the substrate W and the decoder F.sub.2 simulates scattering of radiation from the structure and detection of the detected representation by the metrology apparatus 30.sub.1-301.sub.N based on the geometrical model of the structure and a metrology recipe defining settings of the metrology apparatus 30.sub.1-301.sub.N. Thus, the encoded representation z may comprise reconstructed geometrical dimensions of the geometrical model, e.g., critical dimension, side wall angle, overlay, etc. In such an embodiment, the training of the first machine learning process 41 may comprise adjusting parameters (e.g., material parameters, nominal stack dimensions, fix/float, etc.) defining the geometrical model (i.e. the geometrical model is parametrized by θ.sub.1) and/or adjusting one or more parameters defining the metrology recipe (i.e. the metrology recipe is parametrized by θ.sub.1). Also see previous citation to [0066].}; and adjust one or more parameters of the ML model based on the comparison to influence a cost function used to train the ML model to improve CD matching between the first and the second SEM data sets, the cost function being a function of the first CD values and the second CD values {adjusting parameters to influence cost function described in [0055]: The output 46 from the training process provides an encoder F.sub.1 and/or decoder F.sub.2 that can process detected representations obtained from different metrology apparatuses 30.sub.1-30.sub.N with an optimized balance between fidelity and confusion, where fidelity represents the extent to which the output y retains information about the target, and confusion represents the extent to which the output y from different metrology apparatuses 30.sub.1-30.sub.N is indistinguishable. In an embodiment, a maximization of fidelity and confusion is achieved by optimizing a cost function 43. An example mathematical form of a suitable cost function 43 is described below with reference to the embodiment of FIG. 6.}.
Claim 9
Mossavat further discloses: wherein the instructions are further configured to cause the computer system to: capture metrology data, obtained via the second SEM system, of another patterned substrate; and convert, via the trained ML model, the captured metrology data into converted metrology data, the converted metrology data of the other patterned substrate having characteristics as if captured by the first SEM system {capture metrology data, convert via trained ML model described in [0079].}.
Claim 10
Mossavat further discloses: wherein the instructions are further configured to cause the computer system to: determine a metrology measurement recipe for the second SEM system based on the first SEM data set and physical characteristic measurements of the patterned substrate from the first SEM system {determine a metrology measurement recipe and claimed features described in [0063]}; capture metrology data of the patterned substrate obtained using the second SEM system {capture described in [0047]}; convert the captured metrology data using the trained machine learning model {convert the captured metrology data using the trained ML model described in [0079]}; and apply the metrology measurement recipe to the converted metrology data to determine another physical characteristic measurement {See previous citation to [0063].}.
Claim 11
Mossavat further discloses: wherein the physical characteristic measurement comprises at least one selected from: a critical dimension (CD) measurement, an overlay measurement, and/or an edge placement error {critical dimension (CD) measurement described in [0029]}.
Claim 12
Mossavat further discloses: wherein the physical characteristic measurement comprising a CD measurement and wherein the metrology measurement recipe comprises one or more CD thresholding values indicative of one or more locations on the captured metrology data where one or more CD measurements be taken {See previous citation to [0029].}.
Claim 16
Mossavat discloses:
A non-transitory computer-readable medium having instructions recorded therein or thereon, the instructions, when executed by a computer system {See [0083].}, configured to cause the computer system to at least:
capture metrology data, obtained by a first metrology system, of a patterned substrate {capture metrology data, obtained by a first metrology system, of a patterned substrate described in [0047]: In an embodiment, the training data 32 is obtained by measuring the same substrate W or set of substrates W with all of the multiple metrology apparatuses 30.sub.1-30.sub.N to be matched. In the example shown, the training data 32 is obtained by performing metrology measurements on two different sets 34.sub.1 and 34.sub.2 of substrates. Metrology apparatuses 30.sub.1-30.sub.3 contribute to the training data 32 by measuring substrates W in substrate set 34.sub.1 and metrology apparatuses 30.sub.4-30.sub.N contribute to the training data 32 by measuring substrates W in substrate set 34.sub.2. patterned substrate also described in [0025].}; and
convert the captured metrology data via a trained ML model into converted metrology data, the converted metrology data having characteristics as if captured by a second metrology system {convert the captured metrology data via a trained ML model into converted metrology data, the converted metrology data having characteristics as if captured by a second metrology system described in [0079]: In any of the embodiments discussed above, an initial training of the first machine learning process 41 and the second machine learning process 42 may be performed using detected representations obtained from a set of metrology apparatuses 30.sub.1-30.sub.N over a set of representative applications. In an alternative approach, a selected one metrology apparatus may be used as a reference for training other metrology apparatuses to achieve matching to the reference metrology apparatus. This training may be performed using a transfer-learning technique. In an embodiment, as depicted in FIG. 9, a virtual metrology apparatus 30.sub.V is used as the reference. Training target profiles 50 are provided to the virtual metrology apparatus 30.sub.V. Simulation of measurement of the training target profiles by the virtual metrology apparatus 30.sub.V provides datasets DS to be input to the encoder F.sub.1 to provide encoded representations CD. This approach may improve the accuracy of the metrology apparatus 30. It is desirable that the training target profiles 50 used for simulation, at least for parameters of interest, cover the distribution of the actual substrates W to be measured, since all differences will be detected by the classifier CL and used to train the encoder/decoder combination to remove these differences from the encoded representation.}.
Claim 17
Mossavat further discloses: access a first metrology data set acquired by the first metrology system and a second metrology data set acquired by the second metrology system; and train, using the first and second metrology data sets as training data, a machine learning (ML) model such that the trained ML model is configured to convert a metrology data set acquired by the second metrology system to a converted data set having characteristics comparable to metrology data being by the first metrology system {See previous citations to [0047] and [0079]; also see claims 1 and 16.}.
Claim 18
Mossavat further discloses: wherein the first and second metrology data sets comprise scanning electron metrology (SEM) images of the patterned substrate and wherein the instructions configured to cause the computer system to train the ML model are further configured to cause the computer system to: compare the first set of SEM images acquired by the first metrology system and second set of SEM images acquired by the second metrology system; and adjust one or more parameters of the ML model based on the comparison to influence a cost function used to train the ML model to improve matching between the first set of SEM images and ML-generated images using the second set of SEM images as input to the ML model {See previous citation to [0041], [0042], [0066]; also see claim 2.}.
Claim 19
Mossavat further discloses:
A metrology system {[0098]} comprising:
a detector configured to detect a substrate {detector described in [0032].};
a computer system including one or more processors {a computer system including one or more processors described in [0083].}; and
the medium of claim 16 {See previous citations to [0047] and [0079]; also see claim 16.}.
Claim 20
Mossavat further discloses: wherein the metrology system is a scanning electron microscope {SEM described in [0034].}.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Mossavat in view of Kusnadi (US 20110202898).
Claim 6
Mossavat further discloses: training data; machine learning; ML {See previous citation to [0079].}
Mossavat, while disclosing the features above, doesn’t explicitly disclose, however, Kusnadi, in a similar field of endeavor directed to contour alignment for model calibration, teaches: wherein the instructions configured to cause the computer system to train the model are further configured to cause the computer system to: align a first image set or first contours of the first SEM data set with a design layout image or design contours of a design layout; align a second image set or second contours of the second SEM data set with the design layout image or the design contours of the design layout; and use the aligned first image set or contours and the aligned second image set or contours as data used [for] the model {align images, use as data [for] the model described in [0039]-[0043] and [0044]-[0054].}.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Mossavat to include the features of Kusnadi. Given that Mossavat is directed to calibrating a plurality of metrology apparatuses, one of ordinary skill would have looked to Kusnadi, in order to facilitate model calibration based on contours of printed layout features {[0003] of Kusnadi}.
Claims 7 and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Mossavat in view of Ozcan (US 20220114711).
Claim 7
Mossavat, while disclosing the features above, doesn’t explicitly disclose, however, Ozcan, in a similar field of endeavor directed to deep learning microscopy, teaches: wherein the instructions configured to cause the computer system to train the ML model are further configured to cause the computer system to: compare pixel intensity values from the first image set and the second image set; and adjust one or more parameters of the ML model based on the comparison to influence the cost function used to train the ML model {compare pixel intensity, adjust parameters based on comparison described in [0083].}.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Mossavat to include the features of Ozcan. Given that Mossavat is directed to calibrating a plurality of metrology apparatuses, one of ordinary skill would have looked to Ozcan, in order to facilitate the use of deep learning techniques to improve upon and expand microscopy methods and techniques {[0004] of Ozcan}.
Claim 14
Mossavat, while disclosing the features above, doesn’t explicitly disclose, however, Ozcan, in a similar field of endeavor directed to deep learning microscopy, teaches: wherein the first SEM system is manufactured by a first manufacturer, and the second metrology system is manufactured by a second, different manufacturer {devices of different modalities, i.e., manufacturers, described in [0116]: This deep learning-based fluorescence super-resolution approach improves both the field-of-view (FOV) and imaging throughput of fluorescence microscopy tools and can be used to transform lower-resolution and wide-field images acquired using various imaging modalities and hardware into higher resolution ones.}.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Mossavat to include the features of Ozcan. Given that Mossavat is directed to calibrating a plurality of metrology apparatuses, one of ordinary skill would have looked to Ozcan, in order to facilitate the use of deep learning techniques to improve upon and expand microscopy methods and techniques {[0004] of Ozcan}.
Claim 15
Mossavat, while disclosing the features above, doesn’t explicitly disclose, however, Ozcan, in a similar field of endeavor directed to deep learning microscopy, teaches: wherein the ML model is trained using a generative adversarial network (GAN) architecture, the ML model comprising a generator model and a discriminator model {the ML model trained using a generative adversarial network (GAN) architecture, the ML model comprising a generator model and a discriminator model described in [0065]: [0065] FIG. 36 is a typical plot of the loss functions of the Generator network and the Discriminator network models during the GAN training. The loss functions for the generator (G) and the discriminator (D) quickly converge to an equilibrium stage. The discriminator loss keeps stable while the generator loss slightly decreases, which means the MSE and SSIM losses that take a very small portion of the total generator loss are decreasing. In this competition process between G and D, the network gradually refines the learnt super-resolution image transformation and recovers better spatial details. After 60,000 iterations, the discriminator takes advantage and the generator loss begins to increase, which will lead to a mode collapse in the GAN network. Therefore, the trained model at 50,000 iterations was used as the final testing model.}.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Mossavat to include the features of Ozcan. Given that Mossavat is directed to calibrating a plurality of metrology apparatuses, one of ordinary skill would have looked to Ozcan, in order to facilitate the use of deep learning techniques to improve upon and expand microscopy methods and techniques {[0004] of Ozcan}.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Mossavat in view of Ozcan, further in view of Fang (US 20200018944).
Claim 8
The combination of Mossavat and Ozcan, while teaching the features above, doesn’t explicitly teach, however, Fang, in a similar field of endeavor directed to SEM image enhancement, teaches: wherein the instructions configured to cause the computer system to train the ML model are further configured to cause the computer system to determine intensity values from the first image set and the second image set by: application of a first contour extraction algorithm associated with the first SEM system on the first SEM data set; and application of a second contour extraction algorithm associated with the second SEM system on the second SEM data set {application of contour extraction algorithm(s) described in [0055]-[0060]}.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the combination of Mossavat and Ozcan to include the features of Fang. Given that Mossavat is directed to calibrating a plurality of metrology apparatuses, one of ordinary skill would have looked to Fang, in order to facilitate defect detection and identification while maintaining high throughput and high yield {[0021] of Fang}.
Claim 13 is rejected under 35 U.S.C. 103 as being unpatentable over Mossavat in view of Fang.
Claim 13
Mossavat, while disclosing the features above, doesn’t explicitly disclose, however, Fang, in a similar field of endeavor directed to SEM image enhancement, teaches: wherein the instructions configured to cause the computer system to determine the metrology measurement recipe are further configured to cause the computer system to: extract, via a first contour extraction algorithm, a contour from an image of the first SEM data set; establish a cutline at a location across the contour to measure a CD; and determine, based on a signal along the cutline, a CD threshold value corresponding to the measured CD {extract via extraction algorithm(s) and claimed features described in [0055]-[0060]}.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Mossavat to include the features of Fang. Given that Mossavat is directed to calibrating a plurality of metrology apparatuses, one of ordinary skill would have looked to Fang, in order to facilitate defect detection and identification while maintaining high throughput and high yield {[0021] of Fang}.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
“Deep Learning-Based Autonomous Scanning Electron Microscope” (NPL attached), which teaches: In this paper, we propose and implement a deep learning-based autonomous SEM machine, which assesses image quality and controls parameters autonomously to get high quality sample images just as if human experts do. This world's first autonomous SEM machine may be the first step to bring SEMs, previously used only for advanced researches due to its difficulty in use, into much broader applications such as education, manufacture, and mechanical diagnosis, which are previously meant for optical microscopes.
US 20190348331, which teaches: A method where deviations of a characteristic of an image simulated by two different process models or deviations of the characteristic simulated by a process model and measured by a metrology tool, are used for various purposes such as to reduce the calibration time, improve the accuracy of the model, and improve the overall manufacturing process.
US 20190378012, which teaches: Disclosed is a method of determining a characteristic of interest relating to a structure on a substrate formed by a lithographic process, the method comprising: obtaining an input image of the structure; and using a trained neural network to determine the characteristic of interest from said input image. Also disclosed is a reticle comprising a target forming feature comprising more than two sub-features each having different sensitivities to a characteristic of interest when imaged onto a substrate to form a corresponding target structure on said substrate. Related methods and apparatuses are also described.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOHN SAMUEL WASAFF whose telephone number is (571)270-5091. The examiner can normally be reached Monday through Friday 8:00 am to 6:00 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, SARAH MONFELDT can be reached at (571) 270-1833. 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.
JOHN SAMUEL WASAFF
Primary Examiner
Art Unit 3629
/JOHN S. WASAFF/Primary Examiner, Art Unit 3629
Prosecution Insights