DETAILED ACTION
The United States Patent & Trademark Office appreciates the response filed for the
current application that is submitted on 08/04/2026. The United States Patent & Trademark
Office reviewed the following documents submitted and has made the following comments
below.
Amendment
Applicant submitted amendments on 08/04/2026. The Examiner acknowledges the amendment and has reviewed the claims accordingly.
Priority
Receipt is acknowledged that application is a 371 of PCT/US2022/081840. Priority to US PRO 63/301,444 with a priority date of 10/20/2022 is acknowledged under 35 USC 119(e) and 37 CFR 1.78. Copies of certified papers required by 37 CFR 1.55 have been retrieved.
Information Disclosure Statement
The IDS dated 11/05/2024 have been considered and placed in the application file.
Overview
Claims 1-20 are pending in this application and have been considered below.
Claims 1- 20 are rejected.
Applicant Arguments:
In regards to the argument on Argument 1, Applicant/s state/s “Applicant has amended the Title to be "VISUAL DESCRIPTION NETWORK USING SOFT LOGIC BLOCKS," as suggested by the Examiner. Applicant respectfully requests withdrawal of the objection.” therefore, the objection to the specification should be removed (See remarks, Pg 11, Paragraph 3).
In regards to the argument on Argument 2, Applicant/s state/s “Applicant has amended the character of"?" with a value of "O" to indicate the decreasing value trend illustrated in FIG. 5. Applicant respectfully requests withdrawal of the objection.” therefore, the objection to the drawings should be removed (See remarks, Pg 11, Paragraph 4).
In regards to the argument on Argument 3, Applicant/s state/s “Amendment, Applicant
amended claim 13 to depend on claim 11, thus providing antecedent basis for "the loss function,"
as recited in claim 13. Applicant submits that claim 13 particularly points out and distinctly
claims the subject matter which applicant regards as the invention, in accordance with 35 U.S.C.
§ l 12(b.). Applicant respectfully requests withdrawal of this rejection.” therefore, the rejection of 35 U.S.C. § l 12(b.) should be removed (See remarks, Pg 12, Paragraph 2).
In regards to the argument on Argument 4, Applicant/s state/s “the cited references do not disclose or suggest "a placement neural network" and "a template" including "(I) a backend network" and "(2) the template footprint," "wherein the backend network is different than the placement neural network," as recited in amended claims 1 and 14-15. Therefore, independent claims 1 and 14-15 are patentable over the cited references. The dependent claims, i.e., claims 2-13 and 16-20, incorporate the requirements of the respective independent claims. Accordingly, the dependent claims are likewise patentable. Applicant therefore respectfully requests reconsideration and withdrawal of this rejection.” therefore, the rejection 35 U.S.C. § l 03 should be withdrawn (See remarks, Pg 12, paragraph 7).
Examiner’s Responses:
In response to Argument 1, Applicant’s arguments, see Remarks, filed 08/04/2026, have been considered and are persuasive, therefore the specification objection has been withdrawn.
In response to Argument 2, Applicant’s arguments, see Remarks, filed 08/04/2026, have been considered and are persuasive, therefore the drawing objection has been withdrawn.
In response to Argument 3, Applicant’s arguments, see Remarks, filed 08/04/2026, have been considered and are persuasive, therefore the 112(b) rejection has been withdrawn.
In response to Argument 4, Applicant’s arguments, see Remarks, filed 08/04/2026, with respect to the rejection(s) of claims 1-20 under 35 U.S.C. 103 have been considered but are moot in view of new ground(s) of rejection caused by the amendments. Therefore, the original rejection of Claim 1-20 is withdrawn under 35 U.S.C. 103. However a new grounds of rejection is issued for Claims 1-9 and 14-19 under 35 U.S.C. 103 in view of Davis et al. (US Patent Publication US 2019/0261914 A1, hereafter referred to as Davis) in view of Andel et al (US Patent Publication US 2008/0008383 A1, hereafter referred to as Andel) in further view of Zadeh et al (US Patent Publication US 2022/0121884 A1, hereafter referred to as Zadeh). An additional new grounds of rejection is issued for dependent claims 10-13 and 20 under 35 U.S.C. 103 in view of Davis et al. (US Patent Publication US 2019/0261914 A1, hereafter referred to as Davis) in view of Andel et al (US Patent Publication US 2008/0008383 A1, hereafter referred to as Andel) in further view of Zadeh et al (US Patent Publication US 2022/0121884 A1, hereafter referred to as Zadeh) in further view of Jonnalagadda et al (US Patent Pub 2022/0180173 A1, hereafter referred to as Jonnalagadda).
The Examiner finds that Davis and Andel teach on the amended claim language. Davis teaches patches or image data to align images based on a reference image in ¶0151, ¶0155-¶0157. Andel teaches a backend network and uses a probability value quantifying a likelihood in Fig 3, ¶0028, ¶0034, ¶0055 and ¶0022. Applicant argues “the cited references do not disclose or suggest "a placement neural network" and "a template" including "(I) a backend network" and "(2) the template footprint," "wherein the backend network is different than the placement neural network," The Examiner interprets that Davis and Andel teaches the main concept of a machine learning system for finding objects or patterns in images by combining neural networks with logic-like rules while the newly introduced art, Zadeh, teaches the new limitations in the amended claim language. The Examiner will maintain prior art Davis and Andel and details of the rejection are below.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are:
- “ an output device configured to” in claim 8, “an output device” is defined in the specification as “Output devices 138 may include a display device, which may function as an output device using technologies including liquid crystal displays (LCD), quantum dot display, dot matrix displays, light emitting diode (LED) displays, organic light-emitting diode (OLED) displays, cathode ray tube (CRT) displays, e-ink, or monochrome, color, or any other type of display capable of generating tactile, audio, and/or visual output. In some examples, computing system 100 may include a presence-sensitive display that may serve as a user interface device that operates both as one or more input devices 134 and one or more output devices 138.” in ¶0039.
- “ a symbolizer configured to” in claim 10, “a symbolizer” is defined in the specification as “A symbolizer maps continuous parameters into a set of binomial parameters that are combined with unstructured and/or structured predicates as input to an information- theoretic objective function in order to co-adapt the predicates and the attention to placement.” in ¶0092.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
The claims in this application are given their broadest reasonable interpretation using the
plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification.
Under MPEP 2143.03, "All words in a claim must be considered in judging the patentability of that claim against the prior art." In re Wilson, 424 F.2d 1382, 1385, 165 USPQ 494, 496 (CCPA 1970). As a general matter, the grammar and ordinary meaning of terms as understood by one having ordinary skill in the art used in a claim will dictate whether, and to what extent, the language limits the claim scope. Language that suggests or makes a feature or step optional but does not require that feature or step does not limit the scope of a claim under the broadest reasonable claim interpretation. In addition, when a claim requires selection of an element from a list of alternatives, the prior art teaches the element if one of the alternatives is taught by the prior art. See, e.g., Fresenius USA, Inc. v. Baxter Int’l, Inc., 582 F.3d 1288, 1298, 92 USPQ2d 1163, 1171 (Fed. Cir. 2009).
Claim 5 recite “one or more” then listing “shifting, scaling, or rotating.”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 7 recite “one or more” then listing “shifting, scaling, or rotating.”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 8 recite “one or more” then listing “of an indication of the likelihood that the particular pattern is present in the footprint, a truth value for a presence of the particular pattern in the image data, a location of the footprint in the image data, or an object class represented by the particular pattern”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 10 recite “one or more” then listing “of the probability value or a truth value”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 13 recite “one or more” then listing “of the local placement parameters, the probability value, the transformed footprint, or a truth value for a presence of the particular pattern in the image data”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 15 recite “one or more” then listing “of an indication of the likelihood that the particular pattern is present in the footprint, a truth value for a presence of the particular pattern in the image data, a location of the footprint in the image data, or an object class represented by the particular pattern”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 17 recite “one or more” then listing “shifting, scaling, or rotating”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
Claim 19 recite “one or more” then listing “shifting, scaling, or rotating”. Since “one or more” is disjunctive, any one of the elements found in the prior art is sufficient to reject the claim. While citations have been provided for completeness and rapid prosecution, only one element is required. Because, on balance, it appears the disjunctive interpretation enjoys the most specification support and for that reason the disjunctive interpretation (one of A, B OR C) is being adopted for the purposes of this Office Action. Applicant’s comments and/or amendments relating to this issue are invited to clarify the claim language and the prosecution history.
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 1-9 and 14-19 are rejected under 35 U.S.C. 103 as unpatentable over Davis et al. (US Patent Publication US 2019/0261914 A1, hereafter referred to as Davis) in view of Andel et al (US Patent Publication US 2008/0008383 A1, hereafter referred to as Andel) in further view of Zadeh et al (US Patent Publication US 2022/0121884 A1, hereafter referred to as Zadeh).
Regarding Claim 1, Davis teaches a computing system for object detection (Davis ¶0167, ¶0382-¶0383, ¶0393, and discloses a computer device that performs object detection, in this instance the object is legion), the computing system comprising processing circuitry and a storage device , wherein the processing circuitry has access to the storage device and is configured to execute (Davis ¶0391-¶0392 discloses circuity and storage devices connected to execute the functionality) a machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach) comprising:
a placement neural network configured to generate, based on processing a patch of image data (Davis ¶0198, ¶0385 discloses the images being divided into patches to be processed by the neural network, the instant application describes a placement neural network in ¶0061 as outputting coordinates relative to an interest point, therefore the examiner interprets that the precursor "placement" has no special meaning outside of describing what the neural network is used for) local placement parameters for aligning a footprint in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) to a template footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images)
wherein aligning the footprint (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) in the patch of image data to the template footprint according to the generated local placement parameters (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) creates a transformed footprint (Davis ¶0152, ¶0155-¶0157 discloses a set of transformed images results, i.e., the original reference image, and the rotated/warped counterparts to the other images) ; and
a template (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) comprising:
trained to identify, based on the transformed footprint (Davis ¶0152, ¶0155-¶0157 discloses a set of transformed images results, i.e., the original reference image, and the rotated/warped counterparts to the other images),
specified by the template footprint is present in the footprint (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery)
(2) the template footprint (Davis ¶0181 and ¶0152 discloses automatically producing probability results extracting the key points from images to align the reference image and the other images and the instant application describes the backend network in ¶0079 as determine the features the points, therefore the examiner interprets that the precursor "backend" has no special meaning of describing what the neural network is used for).
Davis does not explicitly disclose (1) a backend network, a probability value quantifying a likelihood.
Andel is in the same field of image analysis for pattern recognition. Further, Andel teaches (1) a backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 discloses an artificial neural network classifier that utilizes backpropagation) a probability value quantifying a likelihood (Andel ¶0022 discloses validation using a confidence value exceeding a threshold to determine if the output is sufficiently confident to output the result).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis by incorporating the multiple neural network architecture to process multiple other images as taught by Andel; to make an invention that can automatically process multiple images of an object to determine the probability of the pattern identification in multiple images; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to process images quickly with a high degree of confidence for accurate recognition as disclosed by Andel in ¶0016.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Davis and Andel in combination do not explicitly disclose specifying a particular pattern, that the particular pattern, wherein the backend network is different than the placement neural network, specifying the particular pattern.
Zadeh is in the same field of image analysis for pattern recognition. Further, Zadeh teaches specifying a particular pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches) that the particular pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches)
wherein the backend network is different than the placement neural network (Zadeh Fig 188 and ¶1816 discloses using multiple models and ¶1544 discloses using two different networks a pattern network and a join network)
specifying the particular pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis in view of Andel by incorporating the multiple neural network architecture to process the specific pattern recognition as taught by Zadeh; to make an invention that can automatically process multiple images of an object to determine the probability of the pattern identification in multiple images using multiple different networks; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to create a more efficient AI method by using a smaller number of training samples to improve the performance in terms of cost, efficiency, size, training time, computing/resource requirements, battery lifetime, flexibility, and detection/recognition/prediction accuracy as disclosed by Zadeh in ¶0189, ¶0192-¶0193.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Regarding Claim 2, Davis in view of Andel in view of Zadeh teaches the system of claim 1, wherein the machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach) comprises:
a semantic logic layer (Davis ¶0399-¶0400 describes the logic applied in each layer of the ANN) configured to apply a logic formula to the probability value to generate a truth value for a presence of the particular pattern(Davis ¶0031, ¶0064, and ¶0373 discloses using a histogram to determine the scale at which the sample differs from a healthy image and providing a probability of diagnosis) in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery). See Claim 1 for rationale, its parent claim.
Regarding Claim 3, Davis in view of Andel in view of Zadeh teaches the system of claim 1, wherein the machine learning system comprises:
a semantic logic layer (Davis ¶0399-¶0400 describes the logic applied in each layer of the ANN) configured to apply a logic formula to the probability value and the local placement parameters (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) to generate a truth value for a presence of the particular pattern in the patch of image data (Davis ¶0031, ¶0064, and ¶0373 discloses using a histogram to determine the scale at which the sample differs from a healthy image and providing a probability of diagnosis n¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery). See Claim 1 for rationale, its parent claim.
Regarding Claim 4, Davis in view of Andel in view of Zadeh teaches the system of claim 1,
wherein the placement neural network comprises a first placement neural network (Davis ¶0198, ¶0385 discloses the images being divided into patches to be processed by the neural network, the instant application describes a placement neural network in ¶0061 as outputting coordinates relative to an interest point, therefore the examiner interprets that the precursor "placement" has no special meaning outside of describing what the neural network is used for) ,
wherein the template comprises a first template (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images),
wherein the backend network comprises a first backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation),
wherein the particular pattern is a particular first pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches) , and
wherein the machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach) comprises:
a second placement neural network configured to process data received via an output channel of the first backend network of the first template (Andel ¶0059-¶0055 discloses a secondary neural network and secondary analysis element that uses algorithms to template match different features of the image based on the features locations based on the output of the previous neural networks) to generate second local placement parameters for aligning a second footprint in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) to a second template footprint (Andel ¶0059-¶0055 discloses a secondary neural network and secondary analysis element that uses algorithms to template match different features of the image based on the features locations based on the output of the previous neural networks); and
a second template comprising:
(1) a second backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation that is a pattern classifier , and
(2) the second template footprint,(Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation that is a pattern classifier) the second template configured to process transformed data comprising the second footprint in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery transformed according to the second local placement parameters (Andel ¶0032 discloses the data being transformed in each layer of the network) to generate a probability value quantifying a likelihood that a particular second pattern is present in the second footprint (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation that is a pattern classifier and the node of the output then is used to represent the probability associated with a class to produce a confidence value). See Claim 1 for rationale, its parent claim.
Regarding Claim 5, Davis in view of Andel in view of Zadeh teaches the system of claim 1, wherein the machine learning system is configured to create the transformed footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and transforms the other images to align with the reference image which results in the original image patch being transformed) by:
applying the local placement parameters to perform operations (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) comprising one or more of shifting, scaling, or rotating the footprint template to identify image data comprising pixels included in the footprint template (Davis ¶0095, ¶0152 discloses rotating the pixels so that the pixels can be filtered to have different polarizations set and using the polarized light from the pixels to better identify features); and
interpolating the identified image data to create the transformed footprint (Davis ¶0036, ¶0152 discloses different transformations of data on the images to serve as the metrics for matching the images to line the images up which results in the original image patch being transformed). See Claim 1 for rationale, its parent claim.
Regarding Claim 6, Davis in view of Andel in view of Zadeh teaches the system of claim 1,
wherein the template further comprises fiducial placement parameters (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) that define a fiducial coordinate system (Davis ¶0051 and ¶0389 discloses the pixel coordinates used to associate with the features in the image data), and
wherein the template footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) indicates spatial relationships among multiple objects (Davis ¶0030, ¶0034, ¶0244 discloses spatial luminance as a representation of the scene) by expressing locations of the multiple objects (Davis ¶0097 discloses determining the location of the object of interest) in terms of the fiducial coordinate system (Davis ¶0051 and ¶0389 discloses the pixel coordinates used to associate with the features in the image data). See Claim 1 for rationale, its parent claim.
Regarding Claim 7, Davis in view of Andel in view of Zadeh teaches the system of claim 6, wherein the machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach) is configured create the transformed footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and transforms the other images to align with the reference image, which results in the original image patch being transformed) in the patch by:
applying the fiducial placement parameters (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) to perform operations comprising one or more of shifting, scaling, or rotating the patch of image data (Davis ¶0152 discloses rotating then points so that they become aligned (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) ) to generate a fiducial frame of the patch of image data (Davis ¶0051 and ¶0389 discloses coordinates of the image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery);
applying the local placement parameters to perform operations (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) comprising one or more of shifting, scaling, or rotating the footprint template (Davis ¶0098 discloses scaling the reference image, ¶0151 discloses the images are scaled and spatially aligned (i.e., registered), so that a consistently-sized and oriented frame of reference characterizes all of the images) to identify image data of the fiducial frame comprising pixels included in the footprint template (Davis ¶0151 discloses the images are scaled and spatially aligned (i.e., registered), so that a consistently-sized and oriented frame of reference characterizes all of the images); and
interpolating the identified image data (Davis ¶0151 discloses using the image data to adjust the images to match the correct alignment) of the fiducial frame to generate the transformed footprint (Davis ¶0151 discloses the images are scaled and spatially aligned (i.e., registered), so that a consistently-sized and oriented frame of reference characterizes all of the images). See Claim 1 for rationale, its parent claim.
Regarding Claim 8, Davis in view of Andel in view of Zadeh teaches the system of claim 1, further comprising:
an output device (Andel ¶0079 and Fig 8, 374 discloses an output device) configured to output one or more of an indication of the likelihood that the particular pattern is present in the footprint (Davis ¶0120, ¶0064, ¶0181, ¶0405 discloses pattern recognition and matching techniques using image features and key points to determine the probability of a disease diagnosis), a truth value for a presence of the particular pattern (Andel ¶0022 discloses validation using a confidence value exceeding a threshold to determine if the output is sufficiently confident to output the result) in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery), a location of the footprint in the patch of image data (Andel ¶0059-¶0055 discloses a secondary neural network and secondary analysis element that uses algorithms to template match different features of the image based on the features locations based on the output of the previous neural networks), or an object class represented by the particular pattern (Davis ¶0038-¶0039 discloses classifying the patterns based on shapes, ¶0181 discloses classifying based on normal of begin, ¶0222 discloses classifying by type). See Claim 1 for rationale, its parent claim.
Regarding Claim 9, Davis in view of Andel in view of Zadeh teaches the system of claim 1,
wherein the backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation) is configured to output one or more output channels associated with respective object classes (Davis ¶0034, ¶0043 discloses outputting channels based on the color classification), and
wherein the backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation) is configured to output an indication of an object class of the object classes, the object class represented by the particular pattern, via the corresponding output channel of the one or more output channels (Davis ¶0034, ¶0040, ¶0043, and ¶0216 discloses an output of channels with two or three representations of the image based on color and luminance which is useful is discerning the patterns of the image to aid in diagnosis). See Claim 1 for rationale, its parent claim.
Regarding Claim 14, Davis teaches a computing system (Davis ¶0167, ¶0382-¶0383, ¶0393, and discloses a computer device that performs object detection, in this instance the object is legion) comprising processing circuitry and a storage device, wherein the processing circuitry has access to the storage device (Davis ¶0027 and Fig 1 discloses a system that includes a processor and memory working in conjunction) and is configured to:
receive a specification for a soft logic block (Davis ¶0399 and ¶0400 discloses each artificial neuron receiving a plurality of inputs and outputting a single output which is calculated using a activation function where the ANN topology is chosen by the designer), wherein the specification defines:
a placement neural network configured to generate, based on processing a patch of image data (Davis ¶0198, ¶0385 discloses the images being divided into patches to be processed by the neural network, the instant application describes a placement neural network in ¶0061 as outputting coordinates relative to an interest point, therefore the examiner interprets that the precursor "placement" has no special meaning outside of describing what the neural network is used for) local placement parameters for aligning a footprint in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) to a template footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images)
wherein aligning the footprint (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) in the patch of image data to the template footprint according to the generated local placement parameters (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) creates a transformed footprint (Davis ¶0152, ¶0155-¶0157 discloses a set of transformed images results, i.e., the original reference image, and the rotated/warped counterparts to the other images); and
a template (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) comprising
trained to identify, based on the transformed footprint (Davis ¶0152, ¶0155-¶0157 discloses a set of transformed images results, i.e., the original reference image, and the rotated/warped counterparts to the other images)
specified by the template footprint is present in the footprint (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery),
(2) the template footprint (Davis ¶0181 and ¶0152 discloses automatically producing probability results extracting the key points from images to align the reference image and the other images and the instant application describes the backend network in ¶0079 as determine the features the points, therefore the examiner interprets that the precursor "backend" has no special meaning of describing what the neural network is used for).
Davis does not explicitly disclose (1) a backend network, a probability value quantifying a likelihood.
Andel is in the same field of image analysis for pattern recognition. Further, Andel teaches (1) a backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation) a probability value quantifying a likelihood (Andel ¶0022 discloses validation using a confidence value exceeding a threshold to determine if the output is sufficiently confident to output the result).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis by incorporating the multiple neural network architecture to process multiple other images as taught by Andel; to make an invention that can automatically process multiple images of a object to determine the probability of the pattern identification in multiple images; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to process images quickly with a high degree of confidence for accurate recognition as disclosed by Andel in ¶0016.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Davis and Andel in combination do not explicitly disclose specifying a particular pattern, that the particular pattern, wherein the backend network is different than the placement neural network, specifying the particular pattern.
Zadeh is in the same field of image analysis for pattern recognition. Further, Zadeh teaches specifying a particular pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches), that the particular pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches) wherein the backend network is different than the placement neural network (Zadeh Fig 188 and ¶1816 discloses using multiple models and ¶1544 discloses using two different networks a pattern network and a join network), specifying the particular pattern (Zadeh Fig 59, Fig 119, Fig 159, ¶0328, ¶1695, discloses a pattern matching module that matches the current data state against the predicate of each rule to determine if the data patten matches).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis in view of Andel by incorporating the multiple neural network architecture to process the specific pattern recognition as taught by Zadeh; to make an invention that can automatically process multiple images of a object to determine the probability of the pattern identification in multiple images using multiple different networks; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to create a more efficient AI method by using a smaller number of training samples to improve the performance in terms of cost, efficiency, size, training time, computing/resource requirements, battery lifetime, flexibility, and detection/recognition/prediction accuracy as disclosed by Zadeh in ¶0189, ¶0192-¶0193.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Regarding Claim 15, Davis teaches a method for detecting an object within image data (Davis ¶0027 ¶0167, ¶0382 discloses method that performs object detection, in this instance the object is legion) , the method performed by a computing system executing a machine learning system (Davis ¶0391 discloses the computing system being used to implement instructions including the machine learning) and comprising:
generating, by a placement neural network of the machine learning system (Davis ¶0391 discloses the computing system being used to implement instructions including the machine learning) processing a patch of image data (Davis ¶0198, ¶0385 discloses the images being divided into patches to be processed by the neural network, the instant application describes a placement neural network in ¶0061 as outputting coordinates relative to an interest point, therefore the examiner interprets that the precursor "placement" has no special meaning outside of describing what the neural network is used for), local placement parameters for aligning a footprint in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) to a template footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images)
wherein aligning the footprint (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) in the patch of image data to the template footprint according to the generated local placement parameters (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) creates a transformed footprint (Davis ¶0152, ¶0155-¶0157 discloses a set of transformed images results, i.e., the original reference image, and the rotated/warped counterparts to the other images);
processing, by a template of the machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach),
and (2) the template footprint (Davis ¶0181 and ¶0152 discloses automatically producing probability results extracting the key points from images to align the reference image and the other images and the instant application describes the backend network in ¶0079 as determine the features the points, therefore the examiner interprets that the precursor "backend" has no special meaning of describing what the neural network is used for), the transformed footprint to generate a probability value quantifying a likelihood that the particular pattern is present in the footprint (Davis ¶0064, ¶0181, discloses a stated probability based on the frequency of occurrence based on the matching with the reference image)
one or more of an indication of the likelihood that the particular pattern is present in the footprint (Davis ¶0120, ¶0064, ¶0181, ¶0405 discloses pattern recognition and matching techniques using image features and key points to determine the probability of a disease diagnosis)
or an object class represented by the particular pattern (Davis ¶0038-¶0039 discloses classifying the patterns based on shapes, ¶0181 discloses classifying based on normal of begin, ¶0222 discloses classifying by type).
Davis does not explicitly disclose comprising (1) a backend network, Outputting, a truth value for a presence of the particular pattern in the image data, a location of the footprint in the patch of image data.
Andel is in the same field of image analysis for pattern recognition. Further, Andel teaches comprising (1) a backend network (Andel Fig 3, ¶0028, ¶0034, ¶0055 disclose an artificial neural network classifier that utilizes backpropagation), outputting (Andel ¶0079 and Fig 8, 374 discloses an output device) a truth value for a presence of the particular pattern in the image data (Andel ¶0022 discloses validation using a confidence value exceeding a threshold to determine if the output is sufficiently confident to output the result), a location of the footprint in the patch of image data (Andel ¶0059-¶0055 discloses a secondary neural network and secondary analysis element that uses algorithms to template match different features of the image based on the features locations based on the output of the previous neural networks).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis by incorporating the multiple neural network architecture to process multiple other images as taught by Andel; to make an invention that can automatically process multiple images of a object to determine the probability of the pattern identification in multiple images; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to process images quickly with a high degree of confidence for accurate recognition as disclosed by Andel in ¶0016.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Davis and Andel in combination do not explicitly disclose wherein the backend network is different than the placement neural network.
Zadeh is in the same field of image analysis for pattern recognition. Further, Zadeh teaches wherein the backend network is different than the placement neural network (Zadeh Fig 188 and ¶1816 discloses using multiple models and ¶1544 discloses using two different networks a pattern network and a join network).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis in view of Andel by incorporating the multiple neural network architecture to process the specific pattern recognition as taught by Zadeh; to make an invention that can automatically process multiple images of a object to determine the probability of the pattern identification in multiple images using multiple different networks; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to create a more efficient AI method by using a smaller number of training samples to improve the performance in terms of cost, efficiency, size, training time, computing/resource requirements, battery lifetime, flexibility, and detection/recognition/prediction accuracy as disclosed by Zadeh in ¶0189, ¶0192-¶0193.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Regarding Claim 16, Davis in view of Andel in view of Zadeh teaches the method of claim 15, further comprising:
applying, by the semantic logic layer (Davis ¶0399-¶0400 describes the logic applied in each layer of the ANN), a logic formula to the probability value to generate a truth value for a presence of the particular pattern in the patch of image data (Davis ¶0031, ¶0064, and ¶0373 discloses using a histogram to determine the scale at which the sample differs from a healthy image and providing a probability of diagnosis). See Claim 15 for rationale, its parent claim.
Regarding Claim 17, Davis in view of Andel in view of Zadeh teaches the method of claim 15, wherein creating the transformed footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and transforms the other images to align with the reference image) comprises:
applying the local placement parameters to perform operations (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) comprising one or more of shifting, scaling, or rotating the footprint template to identify image data comprising pixels included in the footprint template (Davis ¶0095, ¶0152 discloses rotating the pixels so that the pixels can be filtered to have different polarizations set and using the polarized light from the pixels to better identify features); and
interpolating the identified image data to create the transformed footprint (Davis ¶0036, ¶0152 discloses using different transformations of data on the images to serve as the metrics for matching the images to line the images up). See Claim 15 for rationale, its parent claim.
Regarding Claim 18, Davis in view of Andel in view of Zadeh teaches the method of claim 15,
wherein the template further comprises fiducial placement parameters (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) that define a fiducial coordinate system (Davis ¶0051 and ¶0389 discloses the pixel coordinates used to associate with the features in the image data), and
wherein the template footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) indicates spatial relationships among multiple objects (Davis ¶0030, ¶0034, ¶0244 discloses spatial luminance as a representation of the scene) by expressing locations of the multiple objects (Davis ¶0097 discloses determining the location of the object of interest) in terms of the fiducial coordinate system (Davis ¶0051 and ¶0389 discloses the pixel coordinates used to associate with the features in the image data). See Claim 15 for rationale, its parent claim.
Regarding Claim 19, Davis in view of Andel in view of Zadeh teaches the method of claim 18, wherein transforming the footprint in the patch comprises:
applying the fiducial placement parameters (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) to perform operations comprising one or more of shifting, scaling, or rotating the patch of image data (Davis ¶0152 discloses rotating then points so that they become aligned) to generate a fiducial frame of the patch of image data (Davis ¶0051 and ¶0389 discloses coordinates of the image data);
applying the local placement parameters to perform operations (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images) comprising one or more of shifting, scaling, or rotating the footprint template (Davis ¶0098 discloses scaling the reference image, ¶0151 discloses the images are scaled and spatially aligned (i.e., registered), so that a consistently-sized and oriented frame of reference characterizes all of the images) to identify image data of the fiducial frame comprising pixels included in the footprint template (Davis ¶0151 discloses the images are scaled and spatially aligned (i.e., registered), so that a consistently-sized and oriented frame of reference characterizes all of the images); and
interpolating the identified image data (Davis ¶0151 discloses using the image data to adjust the images to match the correct alignment) of the fiducial frame to create the transformed footprint (Davis ¶0151 discloses the images are scaled and spatially aligned (i.e., registered), so that a consistently-sized and oriented frame of reference characterizes all of the images). See Claim 15 for rationale, its parent claim.
Claims 10-13 and 20 are rejected under 35 U.S.C. 103 as unpatentable over Davis in view of Andel in view of Zadeh in further view of Jonnalagadda et al (US Patent Pub 2022/0180173 A1, hereafter referred to as Jonnalagadda).
Regarding Claim 10, Davis in view of Andel in view of Zadeh teaches the system of claim 1, wherein the machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach) further comprises:
a symbolizer configured to map one or more of the probability value or a truth value (Andel ¶0022 discloses validation using a confidence value exceeding a threshold to determine if the output is sufficiently confident to output the result) for a presence of the particular pattern in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery) to a scalar (Davis ¶0064, ¶0365 discloses mapping the ranked list of possible pathologies to a histogram or heat map which are examples of scalars).
Davis in view of Andel in view of Zadeh does not explicitly disclose for use as input to a loss function.
Jonnalagadda is in the same field of pattern recognition in images using machine learning. Further, Jonnalagadda teaches for use as input to a loss function (Jonnalagadda ¶0091-¶0093 discloses using a loss function to train the neural network on ground truth images).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis in view of Andel in view of Zadeh by incorporating the information-theoretic loss function as taught by Jonnalagadda; to make an invention that can use the probability of the pattern identification in multiple images to train the neural network; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to reduce a number of false positive instances as disclosed by Jonnalagadda in ¶0069, ¶0092 and ¶0103.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Regarding Claim 11, Davis in view of Andel in view of Zadeh teaches the system of claim 1, wherein the machine learning system (Davis ¶0029, ¶0170 discloses the system using a machine learning approach) is configured to train the placement neural network and the backend neural network by processing training data comprising one or more input images (Davis ¶0039, ¶0244, ¶0246, ¶0298, and ¶0400 discloses training the networks using reference training data consisting of images).
Davis in view of Andel in view of Zadeh does not explicitly disclose to optimize a loss function.
Jonnalagadda is in the same field of pattern recognition in images using machine learning. Further, Jonnalagadda teaches to optimize a loss function (Jonnalagadda ¶0091-¶093 discloses using a loss function to train the neural network on ground truth images to improve the network).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis in view of Andel in view of Zadeh by incorporating the information-theoretic loss function as taught by Jonnalagadda; to make an invention that can use the probability of the pattern identification in multiple images to train the neural network; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to reduce a number of false positive instances as disclosed by Jonnalagadda in ¶0069, ¶0092 and ¶0103.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Regarding Claim 12, Davis in view of Andel in view Zadeh in view of Jonnalagadda teaches the system of claim 11, wherein the loss function comprises an information-theoretic loss function (Jonnalagadda ¶0091-¶0093 discloses using a loss function to train the neural network on ground truth images, the loss function using the probability that a pixel is a part of the image).
Regarding Claim 13, Davis in view of Andel in view Zadeh in view of Jonnalagadda teaches the system of claim 11, wherein inputs to the loss function (Jonnalagadda ¶0091-¶0093 discloses using a loss function to train the neural network on ground truth images) comprises data indicating one or more of the local placement parameters (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and the other images), the probability value (Davis ¶0031, ¶0064, and ¶0373 discloses using a histogram to determine the scale at which the sample differs from a healthy image and providing a probability of diagnosis), the transformed footprint (Davis ¶0152, ¶0155-¶0157 disclose extracting the key points from images to align the reference image and transforms the other images to align with the reference image), or a truth value for a presence of the particular pattern (Andel ¶0022 discloses validation using a confidence value exceeding a threshold to determine if the output is sufficiently confident to output the result)in the patch of image data (Davis ¶0151, ¶0155-¶0157 disclose aligning the images based on a frame of reference which the examiner is interpreting as the same meaning as a template footprint and ¶0198 discloses the image being divided into patches and the patches being compared to the reference imagery).
Regarding Claim 20, Davis in view of Andel in view of Zadeh teaches the method of claim 15, further comprising:
training the placement neural network and the backend neural network by processing training data comprising one or more input images (Davis ¶0039, ¶0244, ¶0246, ¶0298, and ¶0400 discloses training the networks using reference training data consisting of images).
Davis in view of Andel in view of Zadeh does not explicitly disclose to optimize a loss function.
Jonnalagadda is in the same field of pattern recognition in images using machine learning. Further, Jonnalagadda teaches to optimize a loss function (Jonnalagadda ¶0091-¶093 discloses using a loss function to train the neural network on ground truth images to improve the network).
Therefore, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Davis in view of Andel in view of Zadeh by incorporating the information-theoretic loss function as taught by Jonnalagadda; to make an invention that can use the probability of the pattern identification in multiple images to train the neural network; thus one of ordinary skilled in the art would be motivated to combine the references since there is a need to reduce a number of false positive instances as disclosed by Jonnalagadda in ¶0069, ¶0092 and ¶0103.
Thus, the claimed subject matter would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RACHEL ROBERTS whose telephone number is (571)272-6413. The examiner can normally be reached Monday- Friday 7:30am- 5:00pm.
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, Oneal Mistry can be reached on (313) 446-4912. 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.
/RACHEL L ROBERTS/Examiner, Art Unit 2674
/ONEAL R MISTRY/Supervisory Patent Examiner, Art Unit 2674
Prosecution Insights