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 .
Response to Remark(s)
Applicant's amendment filed January 23th, 2026 have been fully entered and considered. Regarding the arguments to the previous prior art rejections and the 101 rejections, the examiner respectfully finds the arguments to be non-persuasive, see response to remarks section below. Accordingly, this action is made final.
Status of Claims
Claims 1-17 and 20-22 are pending, claims 1, 3-4, 6, 9, 20 and 22 have been amended, claims 18-19 have been cancelled, wherein the independent claims 1, 20 and 22 have been amended. Claims 1-17 and 20-22 remains rejected.
Response to Argument(s)
103 rejections:
In pages 7-8 of the remarks, the Applicants argue that the proposed Xie in view of Park , in combination, does not teach or suggest the features of the independent claims 1, 20 and 22:
For claim 1:
“at least two data representations related by a transformation parameterized by at least one numerical transformation value;
wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation”
For claims 20 and 22:
“wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation” as recited in claim 20.
“wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parametrizing the transformation” as recited in claim 22.
For claim 1, in support of the above arguments, the Applicant assert that the proposed Xie specifically does not teach or suggest the “numerical transformation value parameterizing the transformation,” as according to the algorithm 1 of Xie, its contrastive loss is defined over points in two views: we minimize the distance for matched points, and maximize the distance for unmatched points, and that the distance between matched points is not a numerical value parametrizing a transformation that relates at least two data representation. Therefore, Xie is silent about the discussed features. Park also does not teach this limitation.
For claims 20 and 22, in support of the above arguments, the Applicant assert that the for the reasons that should be appreciated from the foregoing from the remarks for claim 1, no combination of Xie and Park would have disclosed the discussed limitations stated above.
Examiner’s reply:
the examiner respectfully disagrees with the Applicants arguments, the Applicants are respectfully reminded that the claims are construed based on BRI (broadest reasonable interpretation) in light of the specification, that the examiner finds the proposed Xie and Park to cover the BRI scope of the claims. Importantly, the proposed Xie teaches the claimed limitation of “wherein the self-supervised regression loss function encourages the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation”, as previously stated in the rejection mapping, Xies the algorithm 1 teaches wherein the contrastive loss is to encourage the numerical output value of f1 and f2 (the recited one numerical output value and the at least one numerical transformation value) to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par. In further explanation of this mapping, since f1 and f2 are two output features extracted based on two transformation values of T1 and T2, and that since the contrastive loss of algorithm 1 is to update the backbone architecture NN on the matched points therefore, the backbone architecture can compute the correspondence mapping (matches) between the points more accurately (as shown in algorithm 1), therefore, what is the contrastive loss is for but encourage correspondence mapping/matching of the points? Moreover, since the stated last step of algorithm 1 is to update the backbone architecture with the contrastive loss on the matched points, it is understood that the steps of algorithm 1 is being performed iteratively hence, it is also have the updated architecture is based on the previously mapped points and the previously sampled transformations being used as variables/parameters for the computed feature points which are then used in the computation of the contrastive loss, therefore, the next iteration of the steps of algorithm 1 is to match the numerical output values wherein, these numerical output values include values of a transformation value parameterizing the transformation, wherein, emphatically, being variables/parameters of another in the same algorithm indicating its parameterizing the transformation such as already discussed. Similarly, for the limitation of “at least two data representations related by a transformation parameterized by at least one numerical transformation value,” Xie teaches the two data representations in FIG. 2 of T1 and T2 relates to each other through a transformation such as, further shown in algorithm 1 of section 3.2, which is parameterized by transformation numerical value T1 and T2. Therefore, the scope of the discussed limitations fall under the scope of Xie’s teaching. Therefore, the rejection still remain.
Same reason for the independent claims 20 and 22. Therefore, the independent claims 20 and 22’s rejections still remain.
101 rejections:
In pages 9-12 of the remarks, the Applicants argue that the 101 rejected claims cannot be realistically executed in the human mind, according to a memorandum titled “reminders in evaluating subject matter eligibility of claims under 35 U.S.C. 101” issued by the USPTO on August 4th, 2025, which emphasizes on the distinction that claims must actually recited an abstract idea rather than those there merely involve one, such as stated in the Applicants’ remarks in page 9.
In support of the above argument, the Applicants state that:
Claims 1-17 and 20-22 are not ineligible under 101 as “Mental Processes:”
Such as claim 1 recites limitations such as “training an encoder to extract features from sensor data” “encoder” and “extract respective features from the at least two data representations of each training example,” “training the encoder based on a self-supervised regression loss function applied to the training examples” wherein “at least one numerical value is computed from the extracted features, wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation” are not steps of mental processes.
Claims 1-17 and 20-22 are not ineligible under 101 as “Mathematical Concepts”
Such as claim 1 cannot be said to be drawn to mathematical calculation if it’s merely based on or involve mathematical concept and not reciting a mathematical concept. Such as stated in the Applicants’ remarks, in page, 10-11.
Claims 1-17 and 20-22 are eligible under 101 as they recite an improvement to the functioning of a computer, or an improvement to other technology or a technical field
As the Applicants believe and state that such as claim 1 improve the “functioning of a computer, or an improvement to other technology or technical field” as it draws to “a computer implemented method of training an encoder to extract features from sensor data” including “training the encoder based on a self-supervised regression loss function applied to the training examples” wherein “the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation.”
Examiner’s reply:
The examiner respectfully disagrees with the Applicants’ arguments regarding the 101. The examiner interprets the claims based on the 101 requirements and find limitations to fall under mental processes, such as “generating a plurality of training examples,….of a set of sensor data;” “extracts respective features….each training example;” “encourage the at least one numerical output value….match….transformation value;” and some others fall under mathematical operation abstract ideas such as “the at least two data….by at least one numerical transformation value;” and “training the encoder……loss function applied to the training examples” and “and at least one numerical output value is computed…..features, wherein the self-supervised regression loss function…..parameterizing the transformation;” and additional elements such as the recited “an encoder” and “the encoder” recited at high level of generality to perform generic encoder’s functions.
The Applicants are noted that under MPEP 2106.04(a)(2)(III), mental process (thinking) “can be performed in the human mind, or by a human using a pen and paper" to be an abstract idea. CyberSource Corp. v. Retail Decisions, Inc., 654 F.3d 1366, 1372, 99 USPQ2d 1690, 1695 (Fed. Cir. 2011): "methods which can be performed mentally, or which are the equivalent of human mental work, are unpatentable abstract ideas the ‘basic tools of scientific and technological work’ that are open to all." (citing Gottschalk v. Benson, 409 U.S. 63, 175 USPQ 673 [1972]). See also Mayo Collaborative Servs. v. Prometheus Labs. Inc., 566 U.S. 66, 71, 101 USPQ2d 1961, 1965 ("mental processes and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work’" (quoting Benson, 409 U.S. at 67, 175 USPQ at 675).
The courts do not distinguish between mental processes that are performed entirely in the human mind and mental processes that require a human to use a physical aid (e.g., pen and paper or a slide rule) to perform the claim limitation. See, e.g., Benson, 409 U.S. at 67, 65, 175 USPQ at 674-75, 674; Synopsys, Inc. v. Mentor Graphics Corp., 839 F.3d 1138, 1139, 120 USPQ2d 1473, 1474 (Fed. Cir. 2016).
Nor do the courts distinguish between claims that recite mental processes performed by humans and claims that recite mental processes performed on a computer, generic circuit or device, or the likes. See " Versata Dev. Group v. SAP Am., Inc., 793 F.3d 1306, 1335, 115 USPQ2d 1681, 1702 (Fed. Cir. 2015). See also Intellectual Ventures I LLC v. Symantec Corp., 838 F.3d 1307, 1318, 120 USPQ2d 1353, 1360 (Fed. Cir. 2016) (‘with the exception of generic computer-implemented steps, there is nothing in the claims themselves that foreclose them from being performed by a human, mentally or with pen and paper.’’).
Because both product/device and process claims may recite a "mental process", the phrase "mental processes" should be understood as referring to the type of abstract idea, and not to the statutory category of the claim. The courts have identified numerous product claims as reciting mental process-type abstract ideas, for instance the product claims to computer systems and computer-readable media in Versata Dev. Group. v. SAP Am., Inc., 793 F.3d 1306, 115 USPQ2d 1681 (Fed. Cir. 2015).
Therefore, according to the Applicants’ arguments in “A. Claims 1-17 and 20-22 are not ineligible under 101 as “Mental Processes”,” the examiner didn’t characterize “an encoder,” “the encoder” to be mental processes but the limitations of “generating a plurality of training examples,….of a set of sensor data;” “extracts respective features….each training example;” “encourage the at least one numerical output value….match….transformation value;” since these steps are processes under mental process (thinking), the human mind can perform being recited to be a mere steps to be implemented by a generic encoder which as understood as a generic tool additional element recited at high level of generality to perform generic functions. Moreover, the limitation of “generating a plurality of training examples,….of a set of sensor data” is a step, by BRI (broadest reasonable interpretation), that a human can also perform mentally through process of observation and evaluation such as a human mind can observe sensor data (here the data/information is already given as data/information being performed the step on/mental process on) by using pen and paper; “extracts respective features….each training example” is a step, by BRI, a human can perform mentally through process of observation and evaluation such as a human can observe training example and extract features, by using pen and paper; “encourage the at least one numerical output value….match….transformation value” can also be understood, under another BRI scope, to be a mental process a human can perform using a pen and paper to encourage two values to match based on observation and evaluation by implementing some observable data/information and make sure they met a certain observable condition/criteria.
Regarding the Applicants’ argument in part “B. Claims 1-17 and 20-22 are not ineligible under 101 as “Mathematical Concepts”,” the examiner find the limitations of “the at least two data….by at least one numerical transformation value;” and “training the encoder……loss function applied to the training examples” and “and at least one numerical output value is computed…..features, wherein the self-supervised regression loss function…..parameterizing the transformation” to explicitly perform mathematical operations and directly involve mathematical concepts such as a recitation of mathematical relationship that “at least two data representations related by a transformation parameterized by at least one numerical transformation value” includes a transformation of numerical values which is a well-known mathematical operation/function of a mathematical transformation process; and “training the encoder based on a self-supervised regression loss function applied to the training examples” is an explicit recitation of a mathematical function to train some generic encoder, of a generic training step using a well-known regression loss function, which is a well-known routine in the art concerning encoder training based on a generic recited at high level of generality training examples; similarly to the limitation of “and at least one numerical output value is computed…..features, wherein the self-supervised regression loss function…..parameterizing the transformation” which explicitly include mathematical operations. Please note that the encoder here characterized as an additional element, there is no further limiting that the encoder is nothing else but a generic, well-known encoder in the art being recited to be trained using generic, well-known training method without further limiting in what specific structure the encoder is bult of, or any specific details how the encoder function to arrive at such output, the training is based on a generic, well-known self-supervised regression loss function, which an ordinary person of ordinary skill in the art can understood and recognize the term is well-known routine.
Moreover, there is no specific application or indication of an improvement of the encoder or a specific practical use of the claimed limitation. Therefore, the Applicants’ argument in “C. Claims 1-17 and 20-22 are eligible under 101 as they recite an improvement to the functioning of a computer, or an improvement to other technology or a technical field” is not persuasive, there is no improvement to a computer functioning, the claimed limitation doesn’t alter or improve the functionality of a computer, but still using generic computer with generic components of a processor executing the instructions of the limitations stored in a generic memory, etc., neither has any improvement to a specific technical field or technology, there is nowhere indicating a specific technical field or technology in the art, the claimed limitation simply recites “a computer implemented method.” The Applicants also fails to meet the 101, step 2A Prong 2’s requirements that there must be additional elements indicated in the claims that integrates the judicial exceptions into a practical application, the examiner finds the additional elements of “an encoder,” “the encoder” recited at high level of generality, generic encoder, recited at high level of generality to perform generic function, hence, not indicating an integration of the judicial exceptions into a practical application, not being considered significantly more.
Therefore, the 101 rejection remains.
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 1-17 and 20-22 are rejected under 35 U.S.C. 101
Regarding Independent Claim 1 and its dependent claims 2-17,
Step 1 Analysis: Claim 1 is directed to a method/process, which falls within one of the four statutory categories.
Step 2A Prong 1 Analysis: Claim 1 recites, in part:
“generating a plurality of training examples, each training example comprising at least two data representations of a set of sensor data, the at least two data representations related by a transformation parameterized by at least one numerical transformation value; and
Training the encoder based on self-supervised regression loss function applied to the training examples;
extracts respective features from the at least two data representations of each training example, and at least one numerical output value is computed from the extracted features, wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation.”
The limitations as drafted, are processes that, under broadest reasonable interpretation, covers the performance of the limitation in the mind which falls within the “Mental Processes/Mathematical Concept” grouping of abstract ideas. The limitations of:
“generating a plurality of training examples,….of a set of sensor data” is a step, by BRI (broadest reasonable interpretation), that a human can also perform mentally through process of observation and evaluation such as a human mind can observe sensor data (here the data/information is already given as data/information being performed the step on/mental process on) by using pen and paper.
“the at least two data….by at least one numerical transformation value;” and “training the encoder……loss function applied to the training examples” and “and at least one numerical output value is computed…..features, wherein the self-supervised regression loss function…..parameterizing the transformation” are all part of a series of mathematical operation abstract ideas.
“extracts respective features….each training example” is a step, by BRI, a human can perform mentally through process of observation and evaluation such as a human can observe training example and extract features, by using pen and paper;
“is configured to drive the at least one numerical output value….match….transformation value” can also be understood, under another BRI scope, to be a mental process a human can perform using a pen and paper to encourage two values to match based on observation and evaluation.
Accordingly, the claim recites an abstract idea.
Step 2A Prong 2 Analysis: This judicial exception is not integrated into a practical application. particular, the claim recites the following additional element(s) –
“an encoder”
“the encoder”
The additional element …encoder… - recited at a high level of generality (i.e. encoder perform generic encoder function recited at high level of generality without further limiting how the encoder works, in details, to arrive at such output, therefore, the recitation of the claimed limitations are just mere attempt to use generic encoder to perform judicial exceptions/abstract ideas) such that it amounts to no more than mere instructions to apply the exception. Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim as a whole is directed to an abstract idea.. Please see MPEP §2106.04.(d).III.C.
Step 2B Analysis: there are no additional elements, such as for these additional elements as indicated above, that amount to significantly more than the judicial exception. Please see MPEP §2106.05. The claim is directed to an abstract idea.
For all of the foregoing reasons, claim 1 does not comply with the requirements of 35 USC 101.
Accordingly, the dependent claims 2-17 do not provide elements that overcome the deficiencies of the independent claim 1. Moreover, claim 2-6, 8, 11-14 and 16-17, each recites, in part, wherein clauses to give further specification to what the mentioned data/information or the abstract ideas of which they each depends on, hence, still merely data/information gathering insignificant extra-solution activities and mere mental processes and mathematical operations with just further specification to the additional elements and/or the abstract ideas. Claim 7 recites, in part, “comparing a first vector or scalar and a second scalar or vector” is a step that can be understood to be mental process and/or mathematical operation of which a human can also compare values using pen and a paper to do a mathematical operation involving comparing of values; “wherein the first vector or scalar……second data representation” is mathematical concept, mathematical relationships of giving definitions to the mathematical terms mentioned in the claim. Claim 9 recites, in part, “wherein the transformation comprises global rotation and…rotation angle;” is a recitation of mathematical relationship; “where at least one….value is computed as…..” is a computation, mathematical operation abstract idea; “the loss function is configured to drive …..rotation angle” is a step that, under BRI, can be understood to be either mental process or mathematical relationship wherein the human mind can determine a loss function to encourage certain condition using pen and paper, and the loss function here being mathematical term hence given the function to encourage certain condition is a mathematical relationship. Claim 10 recites, in part, “wherein the transformation comprises….local rotation angles” is a further specification to indicate that the transformation further comprises mathematical relationship hence, still merely mathematical operation, mathematical relationship; “wherein each local numerical output value is…..the second vector” is a mathematical relationship abstract idea; “the loss function encourages…rotation angles” is a step that, under BRI, can be understood to be either mental process or mathematical relationship wherein the human mind can determine a loss function to encourage certain condition using pen and paper, and the loss function here being mathematical term hence given the function to encourage certain condition is a mathematical relationship. Claim 15 recites, in part, “a 2D object detector” is a generic 2d object detector recited at high level of generality additional element, to be applied to an image other than the at least two data representations which is a recitation of further specification abstract idea, “determine the local transformations…..the sensor data” can be understood to be, based on BRI, mental process or mathematical concept abstract idea such as a human mind can observe data and determine transformations using pen and paper or under another BRI scope, the step can be understood as using mathematical concepts.
Accordingly, the dependent claims 2-17 are not patent eligible under 101.
Regarding Independent Claim 20 and its dependent claim 21,
The independent claim 20 recites analogous limitations to the independent claim 1 hence, these limitations are analyzed under the same approach to be 101 ineligible and the claim 20 is 101 ineligible, moreover claim 20 recites further generic and insignificant additional elements such as “at least one memory configured to store computer-readable instructions,” “at least one hardware processor coupled to the at least one memory and configured to execute the computer-readable instructions, which upon execution cause the at least one hardware processor to extract features from sensor data” are computer and computer components recited at high level of generality to perform generic functions such as memory storing instructions, processor to execute the instructions of the invention hence. Furthermore, the claim has additional element of “a perception component” which is a generic component recited at high level of generality without further limiting how the component works, in details, to arrive at an outcome; another additional element of “receive an input sensor data representation and extract features therefrom, and the perception component is configured to use the extracted features to interpret the input sensor data representation” includes additional elements of insignificant extra-solution activities of data gathering, data using such as receiving data, extracting data and using data to be applied to generic component hence, are not indicative of an integration of the judicial exceptions into a practical application, nor considered significantly more.
The dependent claim 21 recites in part, “wherein the perception component is configured to perform a regression task on the extracted features” which includes the generic component additional element of the perception component recited at high level of generality such as discussed above in claim 20, to recited a regression task on the features hence, is recited to perform an abstract idea of a mental process or mathematical concept, based on BRI.
Regarding Independent Claim 22
The independent claim 22 recites analogous limitations to the independent claim 1 and claim 20 hence, these limitations are analyzed under the same approach to be 101 ineligible and the claim 22 is 101 ineligible, moreover claim 22 recites further generic and insignificant additional elements such as “a non-transitory medium embodying computer-readable instructions,” “executed on one or more hardware processor” which are generic computer components recited at high level of generality to perform generic functions hence, not indicative of an integration of the judicial exceptions into a practical application nor, considered significantly more.
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.
Claims 1-8, 13-17 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Saining Xie et. al. (“PointContrast: Unsupervised Pre-Training for 3D Point Cloud Understanding, Dec. 2020, Part of the book series: Lecture Notes in Computer Science, Vol. 12348” hereinafter as “Xie”) in view of Minwoo Park et. al. (“US 2023/0214654 A1” hereinafter as “Park”).
Regarding claim 1, Xie discloses a computer implemented method of training an encoder to extract features from sensor data, the method comprising (abstract and section 3.5 disclose straining of an encoder for processing data from sensor): generating a plurality of training examples, each training example comprising at least two data representations of a set of sensor data (FIG. 2 discloses a pre-training of a model [which indicates a training], moreover, the training uses training data/examples generated from sensors correspond to a scene of an environment, moreover, each scene captured at time point includes T1 and T2 data representations [two data representations as claimed], under BRI, covers the same scope of the claimed limitation), the at least two data representations related to a transformation parameterized by at least one numerical transformation value (the two data representations in FIG. 2 of T1 and T2 relates to each other through a transformation such as, further shown in algorithm 1 of section 3.2, which is parameterized by transformation numerical value T1 and T2); and training the encoder based on a self-supervised loss function applied to the training examples (section 3.1 discloses the encoder is trained based on self-supervised training using contrastive loss function [section 3, 1st par.]); wherein the encoder extracts respective features from the at least two data representations of each training example (the network of FIG. 2 includes an encoder network hence, for the mapping, it can be understood the network of section 3.5 performs a function of an encoder as well, by BRI, therefore, the steps of FIG. 2 is the steps of an encoder network wherein each scene is extracted feature f1 and feature f2 from the data representations T1 and T2), and at least one numerical output value is computed from the extracted features (the features are being computed point features such as shown in algorithm 1 through their respective formula in the algorithm, by BRI, covers the scope of the claim), wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim).
However, Xie does not explicitly disclose training the encoder based on a self-supervised regression loss.
In the same field of training an encoder for cloud processing ([0034] and [0026], Park) Park discloses training the encoder based on a self-supervised regression loss (as shown in FIG. 1 as the training can include regression loss and further disclosed in [0006]).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie to perform training of an encoder based on self-supervised loss function, wherein the loss function being a regression loss function as taught by Park to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to train an encoder more effectively (abstract and [0006] of Park).
Regarding claim 2, Xie in view of Park, wherein Xie discloses the method of claim 1, wherein the respective features are respective local features contained in respective feature maps extracted from the at least two data representations (the respective features as discussed above in claim 1 extracted from the at least two data representations, moreover, the features being local features according to Xie’s section 3, 1st par. of the respective feature maps as shown in FIG. 1).
Regarding claim 3, Xie in view of Park, wherein Xie discloses the method of claim 2, wherein the transformation comprises a global transformation and the at least one numerical transformation value comprises a global transformation value (the transformation as discussed above in claim 1, which includes transformation on features of global representations such as disclosed in section 3.1, hence, it can be understood as the transformation being global transformation and include global transformation value such as the value computed in algorithm 1, by BRI, covers the scope of the claim), wherein multiple numerical output values are computed from the extracted local features (the multiple numerical output values of algorithms 1 are being computed from features as shown in algorithm 1 being local features such as disclosed in section 3, 1st par.), and the loss function is configured to drive each of the multiple numerical output values to match the global transformation value (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim).
Regarding claim 4, Xie in view of Park, wherein Xie discloses the method of claim 2, wherein the transformation comprises one or more local transformations and the at least one numerical transformation value comprises one or more local transformation values (the transformation as discussed above in claim 1, which includes transformation on features of local features such as disclosed in section 3, 1st par., hence, it can be understood as the transformation being local transformation and include local transformation value such as the value computed in algorithm 1, by BRI, covers the scope of the claim), wherein multiple local numerical output values are computed from the extracted local features (the multiple numerical output values of algorithms 1 are being computed from features as shown in algorithm 1 being local features such as disclosed in section 3, 1st par.), and the loss function is configured to drive each of the local numerical output values to match a corresponding one of the local transformation values (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim).
Regarding claim 5, Xie in view of Park, wherein Xie discloses the method of claim 4, wherein each local numerical output value is determined based on a mapping between a spatial location of a first of the data representations and a second spatial location of a second of the data representations (the local features of section 3, 1st par., as part of the computation of algorithm 1, such as discussed above in claim 4, being computed correspondence mapping between the points of the two data representations based on spatial location x1 and x2 location such as shown in algorithm 1, therefore, by BRI, the teaching covers the same BRI scope of the claimed limitation).
Regarding claim 6, Xie in view of Park, wherein Xie discloses the method of claim 5, wherein the transformation is fully or partially geometric (“or” indicates a selection, therefore, only one of the options is the instant scope of the claim, the examiner selects “partially” for mapping which is a taught in section 3, last 2 pars., wherein the scanning for the processing of algorithm 1 including the transformation being partially scanned indicates partially geometric, by BRI) and the mapping is determined from the transformation (as shown in algorithm 1, the mapping is determined from the transformation, under BRI, covers the scope of the claimed limitation).
Regarding claim 7, Xie in view of Park, wherein Xie discloses the method of claim 5, wherein each local numerical output value is computed by comparing a first vector or scalar and a second scalar or vector (as disclosed in page 578, 1st par, the local numerical output value is computed by feature vectors of the correspondence mapping between the points which indicates comparing of the points for the mapping, by BRI, covers the scope of the claimed limitation; here, the option “vector” is selected instead of the “scalar” for the mapping), wherein the first vector or scalar is defined by the first spatial location and the feature map of the first data representation (the feature vector of page 578, 1st par., being defined by all the information have been discussed above in claim 5 including the spatial location, the feature map such as shown in algorithm 1), and the second vector or scalar is defined by the second spatial location and the feature map of the second data representation (since the feature vector of page 578, 1st par., includes the vectors of both data representation therefore, any of which can be understood to be the 1st vector and the other would be the 2nd vector therefore, by BRI covers the scope of the claimed limitation).
Regarding claim 8, Xie in view of Park, wherein Xie discloses the method of claim 7, wherein the first and second vectors or scalars are computed from the feature maps using a trainable projection component that is trained simultaneously with the encoder (the sparse Res-U-Net of Fig. 2 being trained together with of the same of T1 and T2, hence, it can be understood as both being trained simultaneously, any of which can be understood to be the encoder [as discussed above in claim 1 since its part of an encoder network] and the other would be understood to be a projection component since it can provide the point cloud of f2 or f1 indicating a projection of points into a point cloud map; by BRI, covers the scope of the claimed limitation).
Regarding claim 13, Xie in view of Park, wherein Xie discloses the method of claim 1, wherein the transformation comprises rescaling, translation, cropping and/or tearing as parameterized by parameterized by the at least one numerical transformation value (“or” indicates a selection, the examiner selects “translation” for mapping which is taught in page 580, 1st par.).
Regarding claim 14, Xie in view of Park, wherein Xie discloses the method of claim 1, wherein the transformation comprises at least one non-geometric transformation (the transformation of page 580, 1st par., discloses the transformation comprises non-geometric transformation), such as addition of noise (the training can include natural noise information such as disclosed in page 588, 2nd par., therefore, the transformation here can be understood to base on processing of noise as well which indicates an addition of noise, by BRI, covers the scope of the claimed limitation), that is parameterized by the at least one numerical transformation value (the two data representations in FIG. 2 of T1 and T2 relates to each other through a transformation such as, further shown in algorithm 1 of section 3.2, which is parameterized by transformation numerical value T1 and T2).
Regarding claim 15, Xie in view of Park, wherein Xie discloses the method of claim 4, wherein a 2D object detector is applied to an image other than the at least two data representations in order to determine the local transformations for one or more objects detected in the image (section 3.5 discloses that the 2D ResNet block [2D object detector as claimed, by BRI] is being applied to the image data in order to perform the processing of the algorithm 2 as discussed in claim 4 including the local transformation), the image containing or associated with the sensor data (as shown in FIG. 2).
Regarding claim 16, Xie in view of Park, wherein Xie discloses the method of claim 15, wherein the data representations encode views of the sensor data in a plane other than an image plane of the image (the data representations as discussed in the above claims to encode views of the sensor data in plane such as shown in FIG.2 to be other than an image plane since it’s in 3D point cloud as well).
Regarding claim 17, Xie in view of Park, wherein Xie discloses the method of a claim 1, wherein the data representations are image or voxel representations (“or” indicates a selection, the examiner selects “image” for mapping which is taught in Fig. 2 of Xie) and wherein the data representations are optionally image or voxel representations of 2D or 3D point clouds (“or” indicates a selection, the examiner selects 3D for mapping which is shown in FIG. 2 of the 3D point cloud of Xie; by BRI, covers the scope of the claimed limitation).
Regarding claim 22, Xie discloses a non-transitory medium embodying training computer-readable instructions program configured, when executed on one or more computer hardware processors (abstract discloses using of an encoder for computer processing hence, indicates the use of a computer to have computer components to perform computer component functions such as a processor to execute instructions of the invention stored in a memory), to train an encoder to extract features from sensor data by: (abstract and section 3.5 disclose straining of an encoder for processing data from sensor): generating a plurality of training examples, each training example comprising at least two data representations of a set of sensor data (FIG. 2 discloses a pre-training of a model [which indicates a training], moreover, the training uses training data/examples generated from sensors correspond to a scene of an environment, moreover, each scene captured at time point includes T1 and T2 data representations [two data representations as claimed], under BRI, covers the same scope of the claimed limitation), the at least two data representations related to a transformation parameterized by at least one numerical transformation value (the two data representations in FIG. 2 of T1 and T2 relates to each other through a transformation such as, further shown in algorithm 1 of section 3.2, which is parameterized by transformation numerical value T1 and T2); and training the encoder based on a self-supervised loss function applied to the training examples (section 3.1 discloses the encoder is trained based on self-supervised training using contrastive loss function [section 3, 1st par.]); wherein the encoder extracts respective features from the at least two data representations of each training example (the network of FIG. 2 includes an encoder network hence, for the mapping, it can be understood the network of section 3.5 performs a function of an encoder as well, by BRI, therefore, the steps of FIG. 2 is the steps of an encoder network wherein each scene is extracted feature f1 and feature f2 from the data representations T1 and T2), and at least one numerical output value is computed from the extracted features (the features are being computed point features such as shown in algorithm 1 through their respective formula in the algorithm, by BRI, covers the scope of the claim), wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim).
However, Xie does not explicitly disclose training the encoder based on a self-supervised regression loss.
In the same field of training an encoder for cloud processing ([0034] and [0026], Park) Park discloses training the encoder based on a self-supervised regression loss (as shown in FIG. 1 as the training can include regression loss and further disclosed in [0006]).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie to perform training of an encoder based on self-supervised loss function, wherein the loss function being a regression loss function as taught by Park to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to train an encoder more effectively (abstract and [0006] of Park).
Claims 9-10 are rejected under 35 U.S.C. 103 as being unpatentable over Saining Xie et. al. (“PointContrast: Unsupervised Pre-Training for 3D Point Cloud Understanding, Dec. 2020, Part of the book series: Lecture Notes in Computer Science, Vol. 12348” hereinafter as “Xie”) in view of Minwoo Park et. al. (“US 2023/0214654 A1” hereinafter as “Park”) and Yingzi Ma (“Self-Supervised Learning of 3D Point Clouds via Feature Transformation and Rotation Prediction, Nov. 2022, 2022 International Conference on Electrical, Computer, Communications and Mechatronics Engineering” hereinafter as “Ma”).
Regarding claim 9, Xie in view of Park, wherein Xie discloses the method of claim 7, wherein the transformation comprises global rotation and the at least one at least one numerical transformation value (as discussed above in claim 1 to include the one numerical transformation and global transformation, moreover, the transformation include rotation as disclosed in page 580, 1st par., hence, analogous to global rotation as claimed, by BRI, since as discussed previously) comprises a global rotation angle (including the rotation as discussed previously of page 580, 1st par., a rotation value/data indicates a rotation angle, such as shown in FIG. 2); and the loss function is configured to drive each of the local numerical output values to match the global rotation angle (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim, here the matching would include the information of the transformation rotation hence would include matching value with a global rotation angle, by BRI).
However, Xie in view of Park does not explicitly disclose wherein at least one local numerical output value is computed as an angular separation between the first vector and the second vector.
In the same field of processing 3D point cloud data (title, Ma) Ma discloses wherein at least one local numerical output value is computed as an angular separation between the first vector and the second vector (section III.B discloses the transformation being transformation rotation including an angle as separation between two vectors, hence, can be combined with Xie to teach the claimed limitation, by BRI).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie in view of Park to perform computing of two vectors and having a transformation including rotation angle moreover, wherein at least one local numerical output value is computed as an angular separation between the first vector and the second vector as taught by Ma to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to have a model that perform data processing and point cloud processing more effectively (abstract, Ma).
Regarding claim 10, Xie in view of Park, wherein Xie discloses the method of claim 7, wherein the transformation comprises local rotation and the at least one at least one numerical transformation value (as discussed above in claim 1 to include the one numerical transformation and global transformation, moreover, the transformation include rotation as disclosed in page 580, 1st par., hence, analogous to local rotation as claimed, by BRI) comprises a local rotation angle (including the rotation as discussed previously of page 580, 1st par., a rotation value/data indicates a rotation angle, such as shown in FIG. 2); and the loss function encourages each of the local numerical output values to match the local rotation angle (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim, here the matching would include the information of the transformation rotation hence would include matching value with a local rotation angle, by BRI).
However, Xie in view of Park does not explicitly disclose wherein at least one local numerical output value is computed as an angular separation between the first vector and the second vector.
In the same field of processing 3D point cloud data (title, Ma) Ma discloses wherein at least one local numerical output value is computed as an angular separation between the first vector and the second vector (section III.B discloses the transformation being transformation rotation including an angle as separation between two vectors, hence, can be combined with Xie to teach the claimed limitation, by BRI).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie in view of Park to perform computing of two vectors and having a transformation including rotation angle moreover, wherein at least one local numerical output value is computed as an angular separation between the first vector and the second vector as taught by Ma to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to have a model that perform data processing and point cloud processing more effectively (abstract, Ma).
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Saining Xie et. al. (“PointContrast: Unsupervised Pre-Training for 3D Point Cloud Understanding, Dec. 2020, Part of the book series: Lecture Notes in Computer Science, Vol. 12348” hereinafter as “Xie”) in view of Minwoo Park et. al. (“US 2023/0214654 A1” hereinafter as “Park”) further in view of Qiangeng Xu et. al. (“Grid-GCN for Fast and Scalable Point Cloud Learning, 2020, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5661-5670” hereinafter as “Xu”).
Regarding claim 11, Xie in view of Park discloses the method of claim 7, wherein the mapping (the mapping as discussed above in claim 7).
However, Xie in view of Park does not explicitly disclose the mapping is from a grid cell of the first data representation to a grid cell of the second representation, wherein the first and spatial second locations are grid cell locations.
In the same field of point cloud mapping (title and abstract, Xu) Xu discloses the mapping is from a grid cell of the first data representation to a grid cell of the second representation, wherein the first and spatial second locations are grid cell locations (as shown in FIG. 1 the mapping includes gridConv and the data representation in grid cells therefore, by BRI, covers the scope of the claimed limitation wherein the data representation is in grid cell and the locations are grid cell locations, to be combined with Xie in view of Park to teach the claimed limitation, by BRI).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie in view of Park to perform mapping of data representations based on spatial locations, wherein the mapping is from a grid cell of the first data representation to a grid cell of the second representation, wherein the first and spatial second locations are grid cell locations as taught by Xu to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to perform mapping of point cloud data more effectively (abstract, Xu).
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Saining Xie et. al. (“PointContrast: Unsupervised Pre-Training for 3D Point Cloud Understanding, Dec. 2020, Part of the book series: Lecture Notes in Computer Science, Vol. 12348” hereinafter as “Xie”) in view of Minwoo Park et. al. (“US 2023/0214654 A1” hereinafter as “Park”) further in view of Qiangeng Xu et. al. (“Grid-GCN for Fast and Scalable Point Cloud Learning, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5661-5670” hereinafter as “Xu”) and Michael Goersele et. al. (“Ambient Point Clouds for View Interpolation, July 2010, SIGGRAPH’ 10: ACM SIGGRAPH 2010 papers, article number 95, pp. 1-6” hereinafter as “Goersele”).
Regarding claim 12, Xie in view of Park discloses the method of claim 7, wherein the mapping (the mapping as discussed above in claim 7).
However, Xie in view of Park does not explicitly disclose the mapping is from a grid cell of the first data representation to a region of the second representation spanning multiple grid cells thereof, the second vector or scalar determined via interpolation of vectors of scalars of the multiple grid cells.
In the same field of point cloud mapping (title and abstract, Xu) Xu discloses the mapping is from a grid cell of the first data representation to a region of the second representation spanning multiple grid cells thereof (as shown in FIG. 1 the mapping includes gridConv and the data representation in grid cells therefore, by BRI, covers the scope of the claimed limitation wherein the data representation is in grid cell and the locations are grid cell locations, to be combined with Xie in view of Park to teach the claimed limitation, by BRI; moreover, since a region of a data representation shown in FIG. 1 of Xu such as a box of FIG. 1 would capture grid cells in all directions x, y, x hence can be understood to be analogous to multiple grid cells as claimed, by BRI).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie in view of Park to perform mapping of data representations based on spatial locations, the mapping is from a grid cell of the first data representation to a region of the second representation spanning multiple grid cells thereof as taught by Xu to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to perform mapping of point cloud data more effectively (abstract, Xu).
However, Xie in view of Park and Xu does not explicitly disclose the second vector or scalar determined via interpolation of vectors of scalars of the multiple grid cells.
In the same field of point cloud mapping (title and abstract, Goersele) Goersele discloses the second vector or scalar determined via interpolation of vectors of scalars of the multiple grid cells (the two vectors are being mapped using interpolation of such as shown in FIG. 3 of the multiple points as represented by grid cells as previously mapped to Xu; therefore, by BRI, Goersele can be combined with Xie in view of Park and Xu to each the claimed limitation).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie in view of Park and Xu to perform mapping of data representations based on spatial locations the mapping is from a grid cell of the first data representation to a region of the second representation spanning multiple grid cells thereof, the second vector or scalar determined via interpolation of vectors of scalars of the multiple grid cells as taught by Goersele to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to perform mapping of point cloud data more effectively (abstract, Goersele).
Claims 20-21 are rejected under 35 U.S.C. 103 as being unpatentable over Saining Xie et. al. (“PointContrast: Unsupervised Pre-Training for 3D Point Cloud Understanding, Dec. 2020, Part of the book series: Lecture Notes in Computer Science, Vol. 12348” hereinafter as “Xie”) in view of Minwoo Park et. al. (“US 2023/0214654 A1” hereinafter as “Park”) and Haoming Lu et. al. (“Deep Learning for 3D Point Cloud Understanding: A Survey, Sept. 2020, Computer Vision and Pattern Recognition, Machine Learning” hereinafter as “Lu”).
Regarding claim 20, Xie discloses a computer system comprising: at least one memory configured to store computer-readable instructions; at least one hardware processor coupled to the at least one memory and configured to execute the computer-readable instructions, which upon execution cause the at least one hardware processor to extract features from sensor data, by: (abstract discloses using of an encoder for computer processing hence, indicates the use of a computer to have computer components to perform computer component functions such as a processor to execute instructions of the invention stored in a memory; abstract and section 3.5 disclose straining of an encoder for processing data from sensor): generating a plurality of training examples, each training example comprising at least two data representations of a set of sensor data (FIG. 2 discloses a pre-training of a model [which indicates a training], moreover, the training uses training data/examples generated from sensors correspond to a scene of an environment, moreover, each scene captured at time point includes T1 and T2 data representations [two data representations as claimed], under BRI, covers the same scope of the claimed limitation), the at least two data representations related to a transformation parameterized by at least one numerical transformation value (the two data representations in FIG. 2 of T1 and T2 relates to each other through a transformation such as, further shown in algorithm 1 of section 3.2, which is parameterized by transformation numerical value T1 and T2); and training the encoder based on a self-supervised loss function applied to the training examples (section 3.1 discloses the encoder is trained based on self-supervised training using contrastive loss function [section 3, 1st par.]); wherein the encoder extracts respective features from the at least two data representations of each training example (the network of FIG. 2 includes an encoder network hence, for the mapping, it can be understood the network of section 3.5 performs a function of an encoder as well, by BRI, therefore, the steps of FIG. 2 is the steps of an encoder network wherein each scene is extracted feature f1 and feature f2 from the data representations T1 and T2), and at least one numerical output value is computed from the extracted features (the features are being computed point features such as shown in algorithm 1 through their respective formula in the algorithm, by BRI, covers the scope of the claim), wherein the self-supervised regression loss function is configured to drive the at least one numerical output value to match the at least one numerical transformation value parameterizing the transformation (as shown in algorithm 1, the contrastive loss is to encourage the numerical output value of f1 and f2 to match through a transformation to match the transformed point cloud such as disclosed in page 579, last par.; by BRI, covers the scope of the claim); wherein the encoder is configured to receive an input sensor data representation and extract features therefrom (as shown in FIG. 2 of the encoder network taking in the data representations from sensors and extract features from them).
However, Xie does not explicitly disclose training the encoder based on a self-supervised regression loss; a perception component; and the perception component is configured to use the extracted features to interpret the input sensor data representation.
In the same field of training an encoder for cloud processing ([0034] and [0026], Park) Park discloses training the encoder based on a self-supervised regression loss (as shown in FIG. 1 as the training can include regression loss and further disclosed in [0006]).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie to perform training of an encoder based on self-supervised loss function, wherein the loss function being a regression loss function as taught by Park to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to train an encoder more effectively (abstract and [0006] of Park).
However, Xie in view of Park does not explicitly disclose a perception component; and the perception component is configured to use the extracted features to interpret the input sensor data representation.
In the same field of training a network for 3D point cloud mapping (title and abstract, Lu) Lu discloses a perception component (the features of the point cloud also include perception features such as disclosed in section 5.2.2, 3rd to the last par., therefore, can be understood to have a perception component to it, by BRI); and the perception component is configured to use the extracted features to interpret the input sensor data representation (the perception features are being used to generate conceptual models to enhance feature [section 5.2.2, 3rd to the last par.] to interpret the input sensor data based on BRI, covers the scope of the claim).
Thus, it would have been obvious for a person of ordinary skill in the art before the effective filing date to modify Xie in view of Park to perform training of an encoder based on self-supervised loss function, wherein the loss function being a regression loss function and include a perception component; and the perception component is configured to use the extracted features to interpret the input sensor data representation as taught by Lu to arrive at the claimed invention discussed above. Such a modification is the result of combing prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to learn about 3D point cloud more effectively and enhance understanding of point cloud features (abstract and section 5.2.2, 3rd to the last par., Lu).
Regarding claim 21, Xie in view of Park and Lu discloses the computer system of claim 20, wherein the perception component is configured to perform a regression task on the extracted features (as discussed above in claim 20, Park teaches the regression task, moreover, Lu teaches that the features include perception features, hence, they are combined to teach that the features are being used for the regression task being of a perception component, by BRI, the combination of the discussed proposed arts teach the scope of the claim). The motivation for the combination of the arts is the same as for claim 20 above.
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 PHUONG HAU CAI whose telephone number is (571)272-9424. The examiner can normally be reached M-F 8:30 am - 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, Chineyere Wills-Burns can be reached at (571) 272-9752. 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.
/PHUONG HAU CAI/ Examiner, Art Unit 2673
/CHINEYERE WILLS-BURNS/Supervisory Patent Examiner, Art Unit 2673
Prosecution Insights