Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Acknowledgement is made of Applicant’s claim of the present application being a 371 of PCT International Application No. PCT/CN2022/114976 filed 08/26/2022.
Information Disclosure Statement
The information disclosure statements (“IDS”) filed 01/23/2025, 02/11/2025, and 06/29/2026 have been reviewed and the listed references were noted.
Drawings
The 12-page drawings have been considered and placed in the file.
Status of Claims
Claims 26-45 are pending.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 26, 36, and 43 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 1 of copending Application No. 18/857,053 in view of Ying et al. (“Hierarchical Graph Representation Learning with Differentiable Pooling”).
Copending Application No. 18/857,053 teaches, “A method, the method comprising: receiving, by a neural network, a graph representation of an object, the graph representation comprising a plurality of graph nodes;” (Claim 1 discloses; “A method, the method comprising: identifying a plurality of graph nodes from a graphical representation of an object, each graph node representing a component of the object”) “transforming, by the neural network, the graph representation to a grid representation of the object (Claim 1 discloses; “generating a grid representation of the object by arranging the plurality of graph nodes based on a grid that includes a plurality of elements”) “the grid representation comprising a plurality of grid nodes arranged in a grid structure that comprises a plurality of grid elements,” (Claim 1 discloses; “wherein the plurality of graph nodes is assigned to different ones of the plurality of elements of the grid”) “(Claim 1 discloses; “the neural network comprising a convolutional layer configured to extract features from the grid representation of the object”) “and determining, by the neural network, a condition of the object based on the grid- structured feature map.” (Claim 1 discloses; “and determining a condition of the object based on an output of the neural network.”) Copending Application No. 18/857,053 does not explicitly teach “based on an assignment matrix” and “values of elements in the assignment matrix defining whether to assign any of the plurality of graph nodes to any of the plurality of grid elements”. Since the copending application does not explicitly disclose these limitations, Examiner relies on the teachings of Ying in an analogous field of endeavor. Specifically, Ying teaches, teach “based on an assignment matrix” (Ying, Pg.4, Para. 4 discloses; “Using this construction, the GNNs in DIFFPOOL learn to encode a general pooling strategy that is useful for a large set of training graphs. We first describe how the DIFFPOOL module pools nodes at each layer given an assignment matrix; following this, we discuss how we generate the assignment matrix using a GNN architecture.” And Ying, Pg. 4, Para. 5 discloses; “We denote the learned cluster assignment matrix at layer l as S(l) ∈ Rnl×nl+1. Each row of S(l) corresponds to one of the nl nodes (or clusters) at layer l, and each column of S(l) corresponds to one of the nl+1 clusters at the next layer l + 1.” Ying uses this assignment matrix in their GNNs, which transforms graphs to grids.) and “values of elements in the assignment matrix defining whether to assign any of the plurality of graph nodes to any of the plurality of grid elements” (Ying, Pg. 2, Para. 3 discloses; “Our approach DIFFPOOL learns a differentiable soft assignment at each layer of a deep GNN, mapping nodes to a set of clusters based on their learned embeddings.”)
Copending Application No. 18/857,053 and Ying are considered to be analogous to the claimed invention because they are in the same field of using GNNs to convert 2D representations to 3D representations. Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified copending Application No. 18/857,053 to incorporate the teachings of Ying in order to assign the 2D nodes to grid elements. One of ordinary skill in the art would have been motivated to combine the previously described method of copending Application No. 18/857,053 with the teachings of Ying to increase ability to classify, reduce cost, and ensure differentiability (Ying, Pg. 9, “Conclusion” discloses; “Interesting future directions include learning hard cluster assignments to further reduce computational cost in higher layers while also ensuring differentiability, and applying the hierarchical pooling method to other downstream tasks that require modeling of the entire graph structure.”)
This is a provisional nonstatutory double patenting rejection.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 26, 27, 33, 34, 36, 37, 40, 41, and 43 are rejected under 35 U.S.C. 103 as being unpatentable over Zhao et al. (“Semantic Graph Convolutional Networks for 3D Human Pose Regression”), in view of Ying et al. (“Hierarchical Graph Representation Learning with Differentiable Pooling”).
Regarding claim 26, Zhao teaches, “A method, the method comprising: receiving, by a neural network, a graph representation of an object, the graph representation comprising a plurality of graph nodes;” (Zhao, Pg. 1, Para. 4 discloses; “First, to handle the issue that graph nodes may have various numbers of neighborhoods, the convolution filter shares the same weight matrix for all nodes, which is not comparable with CNNs.”) “(Zhao, Pg. 12, Section A.2. discloses; “To be specific, when the graph convolutional filter is applied to node i in the graph, one matrix W0 is employed to transform the representation of node i while the other matrix W1 is learned for all its neighbors.” Examiner interprets transforming to be transforming the node into a 3D grid representation. Figure 5 also shows the grid representations comprising a plurality of nodes in a 3D grid structure.) “(Zhao, Pg. 5, Para. 2 discloses; “As we demonstrate in experiments, SemGCN manages to effectively encode the mapping from 2D to 3D poses” and Zhao, Pg. 3, Para. 4 discloses; “If we represent the image feature map as a square grid graph whose nodes represent pixels, this approximated formulation can be directly extended to GCNs as shown in Fig.1(c).”) “and determining, by the neural network, a condition of the object based on the grid- structured feature map.” (Zhao, Pg. 8, Para. 2 discloses; “As seen, our method is able to accurately predict 3D pose for both indoor and most in-the-wild images.” And Zhao, Pg. 4, Para. 2 discloses; “At the beginning of the network, one SemGConv is used for mapping the inputs into the latent space; and we have an additional SemGConv which projects the encoded features back to the output space”). Zhao does not explicitly teach, “transforming, by the neural network, the graph representation to a grid representation of the object based on an assignment matrix” and “values of elements in the assignment matrix defining whether to assign any of the plurality of graph nodes to any of the plurality of grid elements”. Since Zhao does not explicitly disclose these limitations, Examiner relies on the teachings of Ying in an analogous field of endeavor. Specifically, Ying teaches, “transforming, by the neural network, the graph representation to a grid representation of the object based on an assignment matrix” (Ying, Pg.4, Para. 4 discloses; “Using this construction, the GNNs in DIFFPOOL learn to encode a general pooling strategy that is useful for a large set of training graphs. We first describe how the DIFFPOOL module pools nodes at each layer given an assignment matrix; following this, we discuss how we generate the assignment matrix using a GNN architecture.” And Ying, Pg. 4, Para. 5 discloses; “We denote the learned cluster assignment matrix at layer l as S(l) ∈ Rnl×nl+1. Each row of S(l) corresponds to one of the nl nodes (or clusters) at layer l, and each column of S(l) corresponds to one of the nl+1 clusters at the next layer l + 1.” Ying uses this assignment matrix in their GNNs, which transforms graphs to grids.) and “values of elements in the assignment matrix defining whether to assign any of the plurality of graph nodes to any of the plurality of grid elements” (Ying, Pg. 2, Para. 3 discloses; “Our approach DIFFPOOL learns a differentiable soft assignment at each layer of a deep GNN, mapping nodes to a set of clusters based on their learned embeddings.”)
Zhao and Ying are considered to be analogous to the claimed invention because they are in the same field of using GNNs to convert 2D representations to 3D representations. Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified Zhao to incorporate the teachings of Ying in order to assign the 2D nodes to grid elements. One of ordinary skill in the art would have been motivated to combine the previously described method of Zhao with the teachings of Ying to increase ability to classify, reduce cost, and ensure differentiability (Ying, Pg. 9, “Conclusion” discloses; “Interesting future directions include learning hard cluster assignments to further reduce computational cost in higher layers while also ensuring differentiability, and applying the hierarchical pooling method to other downstream tasks that require modeling of the entire graph structure.”)
Regarding claim 27, the combination of Zhao and Ying teaches, “The method of claim 26, wherein the values of the elements in the assignment matrix are determined through a process of training the neural network.” (Ying, Pg. 5, “Learning the assignment matrix” discloses; “adjacency matrix A and the node features F for the original graph. At the penultimate layer L − 1 of a deep GNN model using DIFFPOOL, we set the assignment matrix S(L−1) be a vector of 1’s, i.e., all nodes at the final layer L are assigned to a single cluster, generating a final embedding vector corresponding to the entire graph. This final output embedding can then be used as feature input to a differentiable classifier (e.g., a softmax layer), and the entire system can be trained end-to-end using stochastic gradient descent.”) The proposed combination as well as the motivation for combining the Zhao and Ying references presented in the rejection of claim 26, apply to claim 27 and are incorporated herein by reference. Thus, the method recited in claim 27 is met by Zhao and Ying.
Regarding claim 33, the combination of Zhao and Ying teaches, “The method of claim 26, wherein determining a condition of the object based on the grid-structured feature map comprises: determining a pose of the object based on the grid-structured feature map.” (Zhao, Pg. 1, Para. 6 discloses; “In particular, we study SemGCN for 2D to 3D human pose regression. Given a 2D human pose (and the optional relevant image) as input, we aim to predict the locations of its corresponding 3D joints in a certain coordinate space.” And Zhao, Pg. 3, Para. 4 discloses; “If we represent the image feature map as a square grid graph whose nodes represent pixels, this approximated formulation can be directly extended to GCNs as shown in Fig.1(c).”)
Regarding claim 34, the combination of Zhao and Ying teaches, “The method of claim 26, wherein the graph representation is a two-dimensional graph representation, and determining a condition of the object based on the grid-structured feature map comprises: generating a three-dimensional graph representation of the object based on the grid- structured feature map.” (Zhao, Pg. 3, Para. 4 discloses; “If we represent the image feature map as a square grid graph whose nodes represent pixels, this approximated formulation can be directly extended to GCNs as shown in Fig.1(c).” Figure 3 of Zhao also shows the output being a three-dimensional representation of the object. It is inherently based on the feature map since the feature map is part of the process and the 3D representation is the output of the process.)
Claim 36 recites a system with elements corresponding to the steps recited in Claim 26. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao and Ying references, presented in rejection of Claim 26, apply to this claim. Finally, the combination of Zhao and Ying references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claim 37 recites a system with elements corresponding to the steps recited in Claim 27. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao and Ying references, presented in rejection of Claim 27, apply to this claim. Finally, the combination of Zhao and Ying references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claim 40 recites a system with elements corresponding to the steps recited in Claim 33. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao and Ying references, presented in rejection of Claim 26, apply to this claim. Finally, the combination of Zhao and Ying references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claim 41 recites a system with elements corresponding to the steps recited in Claim 34. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao and Ying references, presented in rejection of Claim 26, apply to this claim. Finally, the combination of Zhao and Ying references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claim 43 recites a system with elements corresponding to the steps recited in Claim 26. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao and Ying references, presented in rejection of Claim 26, apply to this claim. Finally, the combination of Zhao and Ying references discloses a processor and a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claims 28, 38, and 44 are rejected under 35 U.S.C. 103 as being unpatentable over Zhao et al. (“Semantic Graph Convolutional Networks for 3D Human Pose Regression”), in view of Ying et al. (“Hierarchical Graph Representation Learning with Differentiable Pooling”), in further view of Jiang et al. (“Convolutional Neural Nets in Chemical Engineering: Foundations, Computations, and Applications”).
Regarding claim 28, the combination of Zhao and Ying does not explicitly teach, “The method of claim 27, wherein the convolutional operation is performed on the grid representation and a convolutional filter, and values of elements in the convolutional filter are determined in the process of training the neural network.” Since the combination of Zhao and Ying does not explicitly disclose these limitations, Examiner relies on the teachings of Jiang in an analogous field of endeavor. Specifically, Jiang teaches, “The method of claim 27, wherein the convolutional operation is performed on the grid representation and a convolutional filter” (Jiang, Pg. 1, “Introduction” discloses; “Convolutions are applied to the input data with different operators (also known as filters) that seek to extract different types of features (e.g., patterns, gradients, geometrical features).” It would be obvious to use this process with the representations of Zhao.) “and values of elements in the convolutional filter are determined in the process of training the neural network.” (Jiang, Pg. 2, Para. 1 discloses; “The goal of the CNN is to learn optimal operators (and associated features) that best map the input data to the output data.”)
Zhao, Ying, and Jiang are considered to be analogous to the claimed invention because they are in the same field of using neural networks to analyze images. Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Zhao and Ying to incorporate the teachings of Jiang in order to have the convolutional operation applied to the representation and filter. One of ordinary skill in the art would have been motivated to combine the previously described method of Zhao and Ying with the teachings of Jiang to learn patterns and filter values to get the best label (pose) (Jiang, Pg. 2, Para. 1 discloses; “For instance, in recognizing an image (the input), the CNN seeks to learn the patterns of the image that best explains a label given to an image (the output).”)
Claim 38 recites a system with elements corresponding to the steps recited in Claim 28. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao, Ying, and Jiang references, presented in rejection of Claim 28, apply to this claim. Finally, the combination of Zhao, Ying, and Jiang references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claim 44 recites a system with elements corresponding to the steps recited in Claims 27 and 28. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claims. Additionally, the rationale and motivation to combine the Zhao, Ying, and Jiang references, presented in rejection of Claim 28, apply to this claim. Finally, the combination of Zhao, Ying, and Jiang references discloses a processor and a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claims 29-32 and 39 are rejected under 35 U.S.C. 103 as being unpatentable over Zhao et al. (“Semantic Graph Convolutional Networks for 3D Human Pose Regression”), in view of Ying et al. (“Hierarchical Graph Representation Learning with Differentiable Pooling”), in further view of Jang et al. (“Categorical Reparameterization with Gumbel-Softmax”).
Regarding claim 29, the combination of Zhao and Ying does not explicitly teach, “The method of claim 27, wherein the values of the elements in the assignment matrix are determined by: training a learnable matrix in the neural network in the process of training the neural network, values of elements in the learnable matrix having a continuous distribution; and after training the learnable matrix, converting the learnable matrix to the assignment matrix through a discretization operation, the values of the elements in the assignment matrix having a discrete distribution.” Since the combination of Zhao and Ying does not explicitly disclose these limitations, Examiner relies on the teachings of Jang in an analogous field of endeavor. Specifically, Jang teaches, “The method of claim 27, wherein the values of the elements in the assignment matrix are determined by: training a learnable matrix in the neural network in the process of training the neural network, values of elements in the learnable matrix having a continuous distribution;” (Jang, Pg. 2, Para. 1 discloses; “We begin by defining the Gumbel-Softmax distribution, a continuous distribution over the simplex that can approximate samples from a categorical distribution.” And “Samples drawn from the Gumbel-Softmax distribution are continuous during training, but are discretized to one-hot vectors during evaluation” Examiner interprets the one-hot vector to be a matrix having 1 row or 1 column.) “and after training the learnable matrix, converting the learnable matrix to the assignment matrix through a discretization operation, the values of the elements in the assignment matrix having a discrete distribution.” (Jang, Pg. 6, Para. 2 discloses; “Samples drawn from the Gumbel-Softmax distribution are continuous during training, but are discretized to one-hot vectors during evaluation” Examiner interprets the one-hot vector to be a matrix having 1 row or 1 column.)
Zhao, Ying, and Jang are considered to be analogous to the claimed invention because they are in the same field of using neural networks to analyze images. Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Zhao and Ying to incorporate the teachings of Jang in order to change the distribution of the learnable matrix after completing training. One of ordinary skill in the art would have been motivated to combine the previously described method of Zhao and Ying with the teachings of Jang to learn hidden representations and sequence modeling (Jang, Pg. 3, Section 2.2 discloses; “Continuous relaxations of one-hot vectors are suitable for problems such as learning hidden representations and sequence modeling.”)
Regarding claim 30, the combination of Zhao, Ying, and Jang teaches, “The method of claim 29, wherein training the learnable matrix in the neural network in the process of training the neural network comprises: inputting training samples into the neural network, the training samples comprising graph representations of objects and associated with ground-truth labels indicating ground-truth conditions of objects illustrated in the training samples” (Zhao, Pg. 5, Section 5.1 discloses; “We make use of the ground truth 2D joint locations provided in the dataset to align the 3D and 2D poses following the setting of [75].”) “the neural network processing the training samples and outputting conditions of the objects” (Zhao, Pg. 5, Section 5.1 discloses; “We make use of the ground truth 2D joint locations provided in the dataset to align the 3D and 2D poses following the setting of [75].” This implies the ground truth labels are used in training for outputting poses.) “and updating the values of the elements in the learnable matrix with a continuous gradient to minimize a difference between the ground-truth labels and the conditions of the objects.” (Jang, Pg. 3, Section 31. Discloses; “This reparameterization trick is commonly applied to training variational autoencoders with continuous latent variables using backpropagation”. Jang updates values (reparametrizes) values with a continuous gradient. “To minimize a difference between the ground-truth labels and the conditions of the objects” is considered functional language and does not hold patentable weight.) The proposed combination as well as the motivation for combining the Zhao, Ying, and Jang references presented in the rejection of claim 29, apply to claim 30 and are incorporated herein by reference. Thus, the method recited in claim 30 is met by Zhao, Ying, and Jang.
Regarding claim 31 the combination of Zhao, Ying, and Jang teaches, “The method of claim 29, wherein converting the learnable matrix to the assignment matrix through the discretization operation comprises: identifying an element in a row in the learnable matrix, the element in the row having a value that is higher than values of other elements in the row;” (Jang, Pg. 3, Section 2.2 discloses; “Continuous relaxations of one-hot vectors are suitable for problems such as learning hidden representations and sequence modeling. For scenarios in which we are constrained to sampling discrete values (e.g. from a discrete action space for reinforcement learning, or quantized compression), we discretize y using argmax but use our continuous approximation in the backward pass by approximating ∇θz ≈ ∇θy.” Examiner interprets that the argmax identifies the largest value in the row.) “changing the value of the element in the row to a first value; and changing the value of the other elements in the row to a second value that is different from the first value.” (Jang, Pg. 6, Para. 2 discloses; “Samples drawn from the Gumbel-Softmax distribution are continuous during training, but are discretized to one-hot vectors during evaluation” A one-hot vector is that where one value is 1 and the rest are 0. Therefore, Jang is changing the elements to a first and second value.) The proposed combination as well as the motivation for combining the Zhao, Ying, and Jang references presented in the rejection of claim 29, apply to claim 31 and are incorporated herein by reference. Thus, the method recited in claim 31 is met by Zhao, Ying, and Jang.
Regarding claim 32, the combination of Zhao, Ying, and Jang teaches, “The method of claim 31, wherein the first value is 1, and the second value is 0.” (Jang, Pg. 6, Para. 2 discloses; “Samples drawn from the Gumbel-Softmax distribution are continuous during training, but are discretized to one-hot vectors during evaluation” A one-hot vector is that where one value is 1 and the rest are 0.) The proposed combination as well as the motivation for combining the Zhao, Ying, and Jang references presented in the rejection of claim 29, apply to claim 32 and are incorporated herein by reference. Thus, the method recited in claim 32 is met by Zhao, Ying, and Jang.
Claim 39 recites a system with elements corresponding to the steps recited in Claim 29. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao, Ying, and Jang references, presented in rejection of Claim 29, apply to this claim. Finally, the combination of Zhao, Ying, and Jang references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claims 35, 42, and 45 are rejected under 35 U.S.C. 103 as being unpatentable over Zhao et al. (“Semantic Graph Convolutional Networks for 3D Human Pose Regression”), in view of Ying et al. (“Hierarchical Graph Representation Learning with Differentiable Pooling”), in further view of Yao et al. (US 20250384583 A1 w/ EFD of 05/16/2022 and different group of inventors).
Regarding claim 35, the combination of Zhao and Ying does not explicitly teach, “The method of claim 26, wherein performing the convolutional operation on the grid representation comprises: generating a first grid representation and a second grid representation based on the grid representation, the first grid representation having a different structure from the second grid representation; and generating the grid-structured feature map based on the first grid representation and a second grid representation.” Since the combination of Zhao and Ying does not explicitly disclose these limitations, Examiner relies on the teachings of Yao in an analogous field of endeavor. Specifically, Yao teaches, “The method of claim 26, wherein performing the convolutional operation on the grid representation comprises: generating a first grid representation and a second grid representation based on the grid representation, the first grid representation having a different structure from the second grid representation; and generating the grid-structured feature map based on the first grid representation and a second grid representation.” (Yao, Para. [0095] discloses; “The convolutional layer may generate a first grid representation and a second grid representation based on the grid representation. The first grid representation has a different structure from the second grid representation. The first grid representation or second grid representation may include the graph nodes. The convolutional layer can generate the features based on the first grid representation and a second grid representation.”)
Zhao, Ying, and Yao are considered to be analogous to the claimed invention because they are in the same field of using neural networks to analyze images. Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Zhao and Ying to incorporate the teachings of Yao in order to generate a feature map based on grid representations. One of ordinary skill in the art would have been motivated to combine the previously described method of Zhao and Ying with the teachings of Yao to increase the accuracy of the feature map by taking more representations into account.
Claim 42 recites a system with elements corresponding to the steps recited in Claim 35. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao, Ying, and Yao references, presented in rejection of Claim 35, apply to this claim. Finally, the combination of Zhao, Ying, and Yao references discloses a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Claim 45 recites a system with elements corresponding to the steps recited in Claim 35. Therefore, the recited elements of this claim are mapped to the proposed combination in the same manner as the corresponding steps in its corresponding method claim. Additionally, the rationale and motivation to combine the Zhao, Ying, and Yao references, presented in rejection of Claim 35, apply to this claim. Finally, the combination of Zhao, Ying, and Yao references discloses a processor and a non-transitory computer-readable media (Zhao, Pg. 7, Section 5.4 discloses; “Meanwhile, the runtime of SemGCN reduces 10% compared with [34], which is around 1.8ms for a forward pass on a Titan Xp GPU.” The Titan Xp GPU has built in memory.)
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JUSTIN M. OAKES whose telephone number is (571)272-9379. The examiner can normally be reached 7:30am-5pm.
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, Amandeep Saini can be reached at (571) 272-3382. 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.
/JUSTIN M OAKES/Examiner, Art Unit 2662
/AMANDEEP SAINI/Supervisory Patent Examiner, Art Unit 2662
