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 Amendment
This action is responsive to the preliminary amendments and remarks received 25 April 2024. Claims 1 - 22 are currently pending.
Specification
The lengthy specification has not been checked to the extent necessary to determine the presence of all possible minor errors. Applicant’s cooperation is requested in correcting any errors of which applicant may become aware in the specification.
Claim Objections
Claim 1 is objected to because of the following informalities: Lines 9 - 10 of claim 1 recite, in part, “extracted features; entropy encoding” which appears to contain a grammatical error and/or a minor informality. The Examiner suggests amending the claim to --extracted features; and entropy encoding-- in order to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 2 is objected to because of the following informalities: Line 1 of claim 2 recites, in part, “wherein the extraction the features” which appears to contain a grammatical error, inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --wherein the extracted the features-- in order to maintain consistency with line 6 of claim 1 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 3 is objected to because of the following informalities: Line 9 of claim 3 recites, in part, “the N-ary tree” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --the N-ary tree-structure-- in order to maintain consistency with line 3 of claim 1 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 5 is objected to because of the following informalities: Line 2 of claim 5 recites, in part, “a subset of nodes from the set;” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --a subset of nodes from the set of neighboring nodes;-- in order to maintain consistency with line 5 of claim 1 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 5 is objected to because of the following informalities: Line 3 of claim 5 recites, in part, “nodes within the subset is provided” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --nodes within the subset of nodes is provided-- in order to maintain consistency with line 2 of claim 5 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 6 is objected to because of the following informalities: Line 6 of claim 6 recites, in part, “the N-ary tree,” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --the N-ary tree-structure,-- in order to maintain consistency with line 3 of claim 1 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 8 is objected to because of the following informalities: Lines 9 - 10 of claim 8 recite, in part, “extracted features; entropy decoding” which appears to contain a grammatical error and/or a minor informality. The Examiner suggests amending the claim to --extracted features; and entropy decoding-- in order to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 9 is objected to because of the following informalities: Line 1 of claim 9 recites, in part, “wherein the extraction the features” which appears to contain a grammatical error, inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --wherein the extracted the features-- in order to maintain consistency with line 6 of claim 8 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 10 is objected to because of the following informalities: Lines 5 - 6 of claim 10 recite, in part, “current node; providing the obtained output” which appears to contain a grammatical error and/or a minor informality. The Examiner suggests amending the claim to --current node; and providing the obtained output-- in order to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 11 is objected to because of the following informalities: Line 2 of claim 11 recites, in part, “the N-ary tree” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --the N-ary tree-structure-- in order to maintain consistency with line 3 of claim 8 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 13 is objected to because of the following informalities: Line 2 of claim 13 recites, in part, “a subset of nodes from the set;” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --a subset of nodes from the set of neighboring nodes;-- in order to maintain consistency with line 5 of claim 8 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 13 is objected to because of the following informalities: Line 3 of claim 13 recites, in part, “nodes within the subset is provided” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending the claim to --nodes within the subset of nodes is provided-- in order to maintain consistency with line 2 of claim 13 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 14 is objected to because of the following informalities: Line 6 of claim 14 recites, in part, “the N-ary tree,” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending line 6 of claim 14 to --the N-ary tree-structure,-- in order to maintain consistency with line 3 of claim 8 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 19 is objected to because of the following informalities: Line 2 of claim 19 recites, in part, “applying of a softmax layer and obtaining the estimated probabilities as” which appears to contain a grammatical error, inconsistent claim terminology and/or minor informalities. The Examiner suggests amending the claim to --applying [[of]] a softmax layer and obtaining the estimated probabilities of information associated with the current node as-- in order to maintain consistency with line 8 of claim 8 and to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 21 is objected to because of the following informalities: Lines 10 - 11 of claim 21 recite, in part, “extracted features; entropy encode” which appears to contain a grammatical error and/or a minor informality. The Examiner suggests amending the claim to --extracted features; and entropy encode-- in order to improve the clarity and precision of the claim. Appropriate correction is required.
Claim 22 is objected to because of the following informalities: Lines 10 - 11 of claim 22 recite, in part, “extracted features; entropy decode” which appears to contain a grammatical error and/or a minor informality. The Examiner suggests amending the claim to --extracted features; and entropy decode-- in order to improve the clarity and precision of the claim. Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 6, 10, 11, 14, 17 and 19 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 6 recites the limitation "the input to the neural network" in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim 10 recites the limitation "the obtained output for each neighboring node" in line 6. There is insufficient antecedent basis for this limitation in the claim.
Claim 14 recites the limitation "the input to the neural network" in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim 17 recites the limitation "the information associated with a node" in lines 1 - 2. There is insufficient antecedent basis for this limitation in the claim.
Claim 17 recites the limitation "the occupancy of the node" in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim 19 recites the limitation "the context embedding related to the current node" in lines 5 - 6. There is insufficient antecedent basis for this limitation in the claim.
Claim 11 is also rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, due to being dependent upon a rejected base claim but would be withdrawn from the rejection if its base claim overcomes the rejection.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1 - 22 are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. U.S. Publication No. 2021/0150771 A1 in view of Zhu et al. U.S. Publication No. 2023/0100413 A1.
- With regards to claim 1, Huang et al. disclose a method for entropy encoding data of a three-dimensional point cloud applied to an electronic device, (Huang et al., Abstract, Figs. 1, 2 & 6 - 9, Pg. 1 ¶ 0005 - 0006 and 0018 - 0019, Pg. 2 ¶ 0023, Pg. 4 ¶ 0045 - 0050, Pg. 6 ¶ 0072 - 0075, Pg. 7 ¶ 0079 - 0080 and 0085 - 0087, Pg. 10 ¶ 0106 - 0108, Pg. 12 ¶ 0136, Pg. 16 ¶ 0184 - Pg. 17 ¶ 0188) comprising: for a current node in an N-ary tree-structure representing the three-dimensional point cloud: (Huang et al., Abstract, Figs. 2, 4 & 7, Pg. 1 ¶ 0005 and 0018 - 0019, Pg. 3 ¶ 0030, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0047, Pg. 4 ¶ 0050 - 0051, Pg. 10 ¶ 0109 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0135, Pg. 13 ¶ 0146 - Pg. 14 ¶ 0152) obtaining a set of neighboring nodes of the current node; (Huang et a., Figs. 2, 4 & 7, Pg. 1 ¶ 0018 - 0019, Pg. 3 ¶ 0033 - 0036, Pg. 4 ¶ 0041 - 0044 and 0050 - 0051, Pg. 5 ¶ 0058 and 0062, Pg. 11 ¶ 0117 - 0123, Pg. 13 ¶ 0138 and 0142 - 0144, Pg. 15 ¶ 0163 - 0164 and 0167) extracting features of the set of the neighboring nodes by applying a neural network including a layer; (Huang et al., Abstract, Figs. 2, 4, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0036 - Pg. 4 ¶ 0045, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0071 - 0073, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0162 - Pg. 15 ¶ 0167) estimating probabilities of information associated with the current node based on the extracted features; (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0037 - 0039, Pg. 4 ¶ 0042 - 0047 and 0050, Pg. 5 ¶ 0057 and 0062, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0155 - 0160, Pg. 15 ¶ 0163 - 0167) entropy encoding the information associated with the current node based on the estimated probabilities. (Huang et al., Abstract, Figs. 2, 6 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 4 ¶ 0045 - 0047, Pg. 5 ¶ 0063, Pg. 6 ¶ 0074 - 0075, Pg. 11 ¶ 0114, Pg. 12 ¶ 0135, Pg. 13 ¶ 0150 - Pg. 14 ¶ 0152, Pg. 15 ¶ 0167 - 0168, Pg. 16 ¶ 0179 - 0180) Huang et al. fail to disclose explicitly a neural network including an attention layer. Pertaining to analogous art, Zhu et al. disclose a method for entropy encoding data applied to an electronic device, (Zhu et al., Abstract, Figs.1, 3 & 8 - 10, Pg. 1 ¶ 0005 - 0007, Pg. 4 ¶ 0066, Pg. 5 ¶ 0073 - Pg. 6 ¶ 0077, Pg. 7 ¶ 0092 - Pg. 8 ¶ 0100, Pg. 9 ¶ 0104 - 0107, Pg. 16 ¶ 0160 - 0164) comprising: for a current node in an N-ary tree-structure: (Zhu et al., Figs. 7A - 9, Pg. 7 ¶ 0087 - 0092, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0122 - Pg. 12 ¶ 0128, Pg. 15 ¶ 0154 - Pg. 16 ¶ 0160) extracting features by applying a neural network including an attention layer; (Zhu et al., Figs. 5A, 5B, 6A & 7B- 9, Pg. 5 ¶ 0069 - 0073, Pg. 8 ¶ 0096, Pg. 9 ¶ 0102 - 0104, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0129, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0160) and entropy encoding the information associated with the current node. (Zhu et al., Figs. 8 & 9, Pg. 5 ¶ 0069 - 0073, Pg. 7 ¶ 0087 - 0091, Pg. 8 ¶ 0094 - 0096, Pg. 9 ¶ 0102 - 0104, Pg. 12 ¶ 0129 - 0130, Pg. 16 ¶ 0155 - 0160) Huang et al. and Zhu et al. are combinable because they are both directed towards data encoding and decoding systems that utilize neural networks to entropy encode and decode input data. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Huang et al. with the teachings of Zhu et al. This modification would have been prompted in order to enhance the base device of Huang et al. with the well-known and applicable technique Zhu et al. applied to a comparable device. Utilizing a neural network including an attention layer to extract features, as taught by Zhu et al., would enhance the base device of Huang et al. by improving its ability to identify and extract the most significant and representative features from input data, the point cloud, thereby enhancing its ability to accurately and efficiently encode and decode input data. Furthermore, this modification would have been prompted by the teachings and suggestions of Huang et al. that their system can utilize multiple multi-layered perceptions during entropy encoding/decoding, that additional information can be used to improve encoding/decoding efficiency and that their system can additionally or alternatively utilize various other machine learning models, such as recurrent neural networks, convolutional neural networks or other forms of neural networks, see at least page 3 paragraphs 0036 - 0037, page 4 paragraph 0048, page 11 paragraph 0124 - page 12 paragraph 0125, page 17 paragraph 0195 and page 18 paragraph 0202 of Huang et al. Moreover, this modification would have been prompted by the teachings and suggestions of Zhu et al. that an improved rate-distortion loss was achieved by utilizing attention layers with a convolutional neural network (CNN) based coding system, see at least page 5 paragraphs 0069 and 0073 and page 12 paragraph 0129 of Zhu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that a neural network including an attention layer would be utilized by the base device of Huang et al. in order to improve its ability to identify and extract the most significant and representative features from input data, the point cloud, and thus enhance its ability to accurately and efficiently encode and decode input data. Therefore, it would have been obvious to combine Huang et al. with Zhu et al. to obtain the invention as specified in claim 1.
- With regards to claim 2, Huang et al. in view of Zhu et al. disclose the method according to claim 1, wherein the extraction the features is based on relative positional information of a neighboring node and the current node within the three-dimensional point cloud as an input. (Huang et al., Fig. 2, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033 - 0035 and 0039, Pg. 4 ¶ 0041 and 0049 - 0050, Pg. 5 ¶ 0058 - 0062, Pg. 6 ¶ 0071, Pg. 11 ¶ 0117 - 0123, Pg. 12 ¶ 0129 - 0131, Pg. 13 ¶ 0143 - 0144 and 0149, Pg. 15 ¶ 0163 - 0167)
- With regards to claim 3, Huang et al. in view of Zhu et al. disclose the method according to claim 2, wherein the extracting the features of the set of the neighboring nodes by applying the neural network comprises: for each neighboring node within the set of neighboring nodes, (Huang et al., Abstract, Figs. 2, 4 & 7, Pg. 1 ¶ 0005 and 0018 - 0019, Pg. 3 ¶ 0030, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0047, Pg. 4 ¶ 0050 - 0051, Pg. 10 ¶ 0109 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0135, Pg. 13 ¶ 0146 - Pg. 14 ¶ 0152) applying a first neural subnetwork to the relative positional information of the respective neighboring node and the current node; (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0044, Pg. 5 ¶ 00057 - 0062, Pg. 6 ¶ 0071 - 0073, Pg. 11 ¶ 0118 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0134, Pg. 13 ¶ 0138 and 0142 - 0150, Pg. 15 ¶ 0163 - 0167, Pg. 16 ¶ 0174 - 0178) providing an obtained output for each neighboring node as an input to another layer, (Huang et al., Figs. 2 & 5, Pg. 3 ¶ 0036 - 0039, Pg. 5 ¶ 0058 - 0062, Pg. 6 ¶ 0072 - 0073, Pg. 11 ¶ 0120, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 13 ¶ 0145 - 0149, Pg. 16 ¶ 0177 - 0178) and wherein an input to the first neural subnetwork includes a level of the current node within the N-ary tree. (Huang et al., Fig. 2, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033, Pg. 4 ¶ 0041, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0120 - 0121, Pg. 12 ¶ 0131, Pg. 13 ¶ 0142 and 0149) Huang et al. fail to disclose explicitly the attention layer. Pertaining to analogous art, Zhu et al. disclose wherein the extracting the features by applying the neural network comprises: providing an obtained output as an input to the attention layer. (Zhu et al., Figs. 5A, 5B, 6A & 7B - 9, Pg. 5 ¶ 0069, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0129, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0158)
- With regards to claim 4, Huang et al. in view of Zhu et al. disclose the method according to claim 1, wherein the extracting the features of the set of the neighboring nodes by applying the neural network comprises applying a second neural subnetwork to output a context embedding into a subsequent layer within the neural network. (Huang et al., Figs. 2, 5 & 7, Pg. 3 ¶ 0036 - Pg. 4 ¶ 0040, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0072 - 0073, Pg. 11 ¶ 0120, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0135, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0162 - Pg. 15 ¶ 0167, Pg. 16 ¶ 0177 - 0178)
- With regards to claim 5, Huang et al. in view of Zhu et al. disclose the method according to claim 1, wherein the extracting the features of the set of neighboring nodes includes selecting a subset of nodes from the set; (Huang et al., Figs. 2 & 4, Pg. 3 ¶ 0033 - 0035, Pg. 5 ¶ 0058 - 0062, Pg. 11 ¶ 0120 - 0123, Pg. 12 ¶ 0129 - 0134, Pg. 13 ¶ 0148 - 0150, Pg. 15 ¶ 0163 - 0167) and wherein information corresponding to nodes within the subset is provided as an input to a subsequent layer within the neural network, (Huang et al., Figs. 2 & 5, Pg. 3 ¶ 0036 - 0039, Pg. 5 ¶ 0058 - 0062, Pg. 6 ¶ 0072 - 0073, Pg. 11 ¶ 0120, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 13 ¶ 0145 - 0149, Pg. 16 ¶ 0177 - 0178) and wherein the selecting the subset of nodes is performed by a k-nearest neighbor algorithm. (Huang et al., Figs. 2 & 4, Pg. 5 ¶ 0058 - 0062, Pg. 12 ¶ 0129 - 0134, Pg. 13 ¶ 0148 - 0150, Pg. 15 ¶ 0163 - 0167)
- With regards to claim 6, Huang et al. in view of Zhu et al. disclose the method according to claim 1, wherein the input to the neural network includes context information of the set of neighboring nodes, (Huang et al., Figs. 2 & 7, Pg. 1 ¶ 0019, Pg. 3 ¶ 0031 and 0033 - 0036, Pg. 4 ¶ 0041, Pg. 6 ¶ 071 - 0072, Pg. 10 ¶ 0111 - Pg. 11 ¶ 0113, Pg. 11 ¶ 0118 - 0124, Pg. 12 ¶ 0129 - 0131, Pg. 13 ¶ 0142 - 0149) wherein the context information for a node includes one or more of a location of the node, octant information, a depth in the N-ary tree, an occupancy code of a respective parent node, and an occupancy pattern of a subset of nodes spatially neighboring the node. (Huang et al., Fig. 2, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033 - 0035, Pg. 4 ¶ 0041, Pg. 6 ¶ 0070 - 0071, Pg. 11 ¶ 0118 - 0123, Pg. 12 ¶ 0129 - 0131, Pg. 13 ¶ 0142 - 0144)
- With regards to claim 7, Huang et al. in view of Zhu et al. disclose the method according to claim 1. Huang et al. fail to disclose explicitly wherein the attention layer in the neural network is a self-attention layer or wherein the attention layer in the neural network is a multi-head attention layer. Pertaining to analogous art, Huang et al. in view of Zhu et al. disclose wherein the attention layer in the neural network is a self-attention layer (Zhu et al., Fig. 6A, Pg. 5 ¶ 0072 - 0073, Pg. 11 ¶ 0119, Pg. 11 ¶ 0122 - Pg. 12 ¶ 0132, Pg. 16 ¶ 0156) or wherein the attention layer in the neural network is a multi-head attention layer. (Zhu et al., Fig. 6A, Pg. 5 ¶ 0072 - 0073, Pg. 11 ¶ 0119, Pg. 12 ¶ 0124)
- With regards to claim 8, Huang et al. a method for entropy decoding data of a three-dimensional point cloud applied to an electronic device, (Huang et al., Abstract, Figs. 1, 2 & 6 - 9, Pg. 1 ¶ 0005 - 0006 and 0018 - 0019, Pg. 2 ¶ 0024, Pg. 4 ¶ 0045 - 0050, Pg. 5 ¶ 0063, Pg. 6 ¶ 0074 - 0075, Pg. 7 ¶ 0079 - 0080 and 0085 - 0087, Pg. 10 ¶ 0106 - 0108, Pg. 12 ¶ 0136, Pg. 14 ¶ 0152, Pg. 16 ¶ 0181 - Pg. 17 ¶ 0188) the method comprising: for a current node in an N-ary tree-structure representing the three-dimensional point cloud: (Huang et al., Abstract, Figs. 2, 4 & 7, Pg. 1 ¶ 0005 and 0018 - 0019, Pg. 3 ¶ 0030, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0047, Pg. 4 ¶ 0050 - 0051, Pg. 10 ¶ 0109 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0135, Pg. 13 ¶ 0146 - Pg. 14 ¶ 0152) obtaining a set of neighboring nodes of the current node; (Huang et a., Figs. 2, 4 & 7, Pg. 1 ¶ 0018 - 0019, Pg. 3 ¶ 0033 - 0036, Pg. 4 ¶ 0041 - 0044 and 0050 - 0051, Pg. 5 ¶ 0058 and 0062, Pg. 11 ¶ 0117 - 0123, Pg. 13 ¶ 0138 and 0142 - 0144, Pg. 15 ¶ 0163 - 0164 and 0167) extracting features of the set of the neighboring nodes by applying a neural network including a layer; (Huang et al., Abstract, Figs. 2, 4, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0036 - Pg. 4 ¶ 0045, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0071 - 0073, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0162 - Pg. 15 ¶ 0167) estimating probabilities of information associated with the current node based on the extracted features; (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0037 - 0039, Pg. 4 ¶ 0042 - 0047 and 0050, Pg. 5 ¶ 0057 and 0062, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0155 - 0160, Pg. 15 ¶ 0163 - 0167) entropy decoding the information associated with the current node based on the estimated probabilities. (Huang et al., Figs. 2, 6 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 2 ¶ 0024, Pg. 4 ¶ 0047, Pg. 5 ¶ 0057 and 0063, Pg. 6 ¶ 0074 - 0075, Pg. 11 ¶ 0114, Pg. 12 ¶ 0136, Pg. 14 ¶ 0152, Pg. 15 ¶ 0167 - 0168, Pg. 16 ¶ 0179 - 0183) Huang et al. fail to disclose explicitly a neural network including an attention layer. Pertaining to analogous art, Zhu et al. disclose a method for entropy decoding data applied to an electronic device, (Zhu et al., Figs.1, 3 & 8 - 10, Pg. 1 ¶ 0005 - 0007, Pg. 4 ¶ 0066, Pg. 5 ¶ 0073 - Pg. 6 ¶ 0077, Pg. 7 ¶ 0092 - Pg. 8 ¶ 0100, Pg. 9 ¶ 0104 - 0107, Pg. 15 ¶ 0147 - 0151, Pg. 16 ¶ 0160 - 0164) the method comprising: for a current node in an N-ary tree-structure: (Zhu et al., Figs. 7A - 9, Pg. 7 ¶ 0087 - 0092, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0122 - Pg. 12 ¶ 0128, Pg. 15 ¶ 0154 - Pg. 16 ¶ 0160) extracting features by applying a neural network including an attention layer; (Zhu et al., Figs. 5A, 5B, 6A & 7B- 9, Pg. 5 ¶ 0069 - 0073, Pg. 8 ¶ 0096, Pg. 9 ¶ 0102 - 0104, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0129, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0160) and entropy decoding the information associated with the current node based on the estimated probabilities. (Zhu et al., Figs. 8 & 9, Pg. 5 ¶ 0069 - 0073, Pg. 7 ¶ 0087 - 0091, Pg. 8 ¶ 0094 - 0096, Pg. 9 ¶ 0102 - 0105, Pg. 11 ¶ 0120, Pg. 12 ¶ 0129 - 0130 and 0133, Pg. 15 ¶ 0147 - 0151, Pg. 16 ¶ 0158 - 0160) Huang et al. and Zhu et al. are combinable because they are both directed towards data encoding and decoding systems that utilize neural networks to entropy encode and decode input data. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Huang et al. with the teachings of Zhu et al. This modification would have been prompted in order to enhance the base device of Huang et al. with the well-known and applicable technique Zhu et al. applied to a comparable device. Utilizing a neural network including an attention layer to extract features, as taught by Zhu et al., would enhance the base device of Huang et al. by improving its ability to identify and extract the most significant and representative features from input data, the point cloud, thereby enhancing its ability to accurately and efficiently encode and decode input data. Furthermore, this modification would have been prompted by the teachings and suggestions of Huang et al. that their system can utilize multiple multi-layered perceptions during entropy encoding/decoding, that additional information can be used to improve encoding/decoding efficiency and that their system can additionally or alternatively utilize various other machine learning models, such as recurrent neural networks, convolutional neural networks or other forms of neural networks, see at least page 3 paragraphs 0036 - 0037, page 4 paragraph 0048, page 11 paragraph 0124 - page 12 paragraph 0125, page 17 paragraph 0195 and page 18 paragraph 0202 of Huang et al. Moreover, this modification would have been prompted by the teachings and suggestions of Zhu et al. that an improved rate-distortion loss was achieved by utilizing attention layers with a convolutional neural network (CNN) based coding system, see at least page 5 paragraphs 0069 and 0073 and page 12 paragraph 0129 of Zhu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that a neural network including an attention layer would be utilized by the base device of Huang et al. in order to improve its ability to identify and extract the most significant and representative features from input data, the point cloud, and thus enhance its ability to accurately and efficiently encode and decode input data. Therefore, it would have been obvious to combine Huang et al. with Zhu et al. to obtain the invention as specified in claim 8.
- With regards to claim 9, Huang et al. in view of Zhu et al. disclose the method according to claim 8, wherein the extraction the features of the set of the neighboring nodes is based on relative positional information of a neighboring node and the current node within the three-dimensional point cloud as an input. (Huang et al., Fig. 2, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033 - 0035 and 0039, Pg. 4 ¶ 0041 and 0049 - 0050, Pg. 5 ¶ 0058 - 0062, Pg. 6 ¶ 0071, Pg. 11 ¶ 0117 - 0123, Pg. 12 ¶ 0129 - 0131, Pg. 13 ¶ 0143 - 0144 and 0149, Pg. 15 ¶ 0163 - 0167)
- With regards to claim 10, Huang et al. in view of Zhu et al. disclose the method according to claim 9, wherein the extracting the features of the set of the neighboring nodes by applying the neural network comprises: for each neighboring node within the set of neighboring nodes, (Huang et al., Abstract, Figs. 2, 4 & 7, Pg. 1 ¶ 0005 and 0018 - 0019, Pg. 3 ¶ 0030, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0047, Pg. 4 ¶ 0050 - 0051, Pg. 10 ¶ 0109 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0135, Pg. 13 ¶ 0146 - Pg. 14 ¶ 0152) applying a first neural subnetwork to the relative positional information of the respective neighboring node and the current node; (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0044, Pg. 5 ¶ 00057 - 0062, Pg. 6 ¶ 0071 - 0073, Pg. 11 ¶ 0118 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0134, Pg. 13 ¶ 0138 and 0142 - 0150, Pg. 15 ¶ 0163 - 0167, Pg. 16 ¶ 0174 - 0178) providing the obtained output for each neighboring node as an input to another layer. (Huang et al., Figs. 2 & 5, Pg. 3 ¶ 0036 - 0039, Pg. 5 ¶ 0058 - 0062, Pg. 6 ¶ 0072 - 0073, Pg. 11 ¶ 0120, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 13 ¶ 0145 - 0149, Pg. 16 ¶ 0177 - 0178) Huang et al. fail to disclose explicitly the attention layer. Pertaining to analogous art, Zhu et al. disclose wherein the extracting the features by applying the neural network comprises: providing the obtained output as an input to the attention layer. (Zhu et al., Figs. 5A, 5B, 6A & 7B - 9, Pg. 5 ¶ 0069, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0129, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0158)
- With regards to claim 11, Huang et al. in view of Zhu et al. disclose the method according to claim 10, wherein an input to the first neural subnetwork includes a level of the current node within the N-ary tree. (Huang et al., Fig. 2, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033, Pg. 4 ¶ 0041, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0120 - 0121, Pg. 12 ¶ 0131, Pg. 13 ¶ 0142 and 0149)
- With regards to claim 12, Huang et al. in view of Zhu et al. disclose the method according to claim 8, wherein the extracting the features of the set of the neighboring nodes by applying the neural network comprises applying a second neural subnetwork to output a context embedding into a subsequent layer within the neural network. (Huang et al., Figs. 2, 5 & 7, Pg. 3 ¶ 0036 - Pg. 4 ¶ 0040, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0072 - 0073, Pg. 11 ¶ 0120, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0135, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0162 - Pg. 15 ¶ 0167, Pg. 16 ¶ 0177 - 0178)
- With regards to claim 13, Huang et al. in view of Zhu et al. disclose the method according to claim 8, wherein the extracting the features of the set of neighboring nodes includes selecting a subset of nodes from the set; (Huang et al., Figs. 2 & 4, Pg. 3 ¶ 0033 - 0035, Pg. 5 ¶ 0058 - 0062, Pg. 11 ¶ 0120 - 0123, Pg. 12 ¶ 0129 - 0134, Pg. 13 ¶ 0148 - 0150, Pg. 15 ¶ 0163 - 0167) and wherein information corresponding to nodes within the subset is provided as an input to a subsequent layer within the neural network, (Huang et al., Figs. 2 & 5, Pg. 3 ¶ 0036 - 0039, Pg. 5 ¶ 0058 - 0062, Pg. 6 ¶ 0072 - 0073, Pg. 11 ¶ 0120, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 13 ¶ 0145 - 0149, Pg. 16 ¶ 0177 - 0178) and wherein the selecting the subset of nodes is performed by a k-nearest neighbor algorithm. (Huang et al., Figs. 2 & 4, Pg. 5 ¶ 0058 - 0062, Pg. 12 ¶ 0129 - 0134, Pg. 13 ¶ 0148 - 0150, Pg. 15 ¶ 0163 - 0167)
- With regards to claim 14, Huang et al. in view of Zhu et al. disclose the method according to claim 8, wherein the input to the neural network includes context information of the set of neighboring nodes, (Huang et al., Figs. 2 & 7, Pg. 1 ¶ 0019, Pg. 3 ¶ 0031 and 0033 - 0036, Pg. 4 ¶ 0041, Pg. 6 ¶ 071 - 0072, Pg. 10 ¶ 0111 - Pg. 11 ¶ 0113, Pg. 11 ¶ 0118 - 0124, Pg. 12 ¶ 0129 - 0131, Pg. 13 ¶ 0142 - 0149) wherein the context information for a node includes one or more of a location of the node, octant information, a depth in the N-ary tree, an occupancy code of a respective parent node, and an occupancy pattern of a subset of nodes spatially neighboring the node. (Huang et al., Fig. 2, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033 - 0035, Pg. 4 ¶ 0041, Pg. 6 ¶ 0070 - 0071, Pg. 11 ¶ 0118 - 0123, Pg. 12 ¶ 0129 - 0131, Pg. 13 ¶ 0142 - 0144)
- With regards to claim 15, Huang et al. in view of Zhu et al. disclose the method according to claim 8. Huang et al. fail to disclose explicitly wherein the attention layer in the neural network is a self-attention layer. Pertaining to analogous art, Zhu et al. disclose wherein the attention layer in the neural network is a self-attention layer. (Zhu et al., Fig. 6A, Pg. 5 ¶ 0072 - 0073, Pg. 11 ¶ 0119, Pg. 11 ¶ 0122 - Pg. 12 ¶ 0132, Pg. 16 ¶ 0156)
- With regards to claim 16, Huang et al. in view of Zhu et al. disclose the method according to claim 8. Huang et al. fail to disclose explicitly wherein the attention layer in the neural network is a multi-head attention layer. Pertaining to analogous art, Zhu et al. disclose wherein the attention layer in the neural network is a multi-head attention layer. (Zhu et al., Fig. 6A, Pg. 5 ¶ 0072 - 0073, Pg. 11 ¶ 0119, Pg. 12 ¶ 0124)
- With regards to claim 17, Huang et al. in view of Zhu et al. disclose the method according to claim 8, wherein the information associated with a node indicates the occupancy code of the node. (Huang et al., Figs. 2 & 7, Pg. 1 ¶ 0019, Pg. 3 ¶ 0033 and 0037, Pg. 4 ¶ 0043 - 0047 and 0049 - 0053, Pg. 5 ¶ 0062, Pg. 10 ¶ 0111 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0118 - 0120, Pg. 12 ¶ 0125 - 0126 and 0128 - 0135, Pg. 13 ¶ 0146 - 0150, Pg. 14 ¶ 0155 - 0156)
- With regards to claim 18, Huang et al. in view of Zhu et al. disclose the method according to claim 8, wherein the neural network includes a third neural subnetwork, (Huang et al., Figs. 2 & 5, Pg. 3 ¶ 0036 - 0039, Pg. 4 ¶ 0041, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0072, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0131 and 0135, Pg. 13 ¶ 0145 - 0146 and 0150, Pg. 16 ¶ 0177 - 0178, Pg. 17 ¶ 00195, Pg. 18 ¶ 0202) the third neural subnetwork performing the estimating of probabilities of information associated with the current node based on the extracted features as an output of a layer. (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0037 - 0039, Pg. 4 ¶ 0042 - 0047 and 0050, Pg. 5 ¶ 0057 and 0062, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0155 - 0160, Pg. 15 ¶ 0163 - 0167) Huang et al. fail to disclose explicitly the attention layer. Pertaining to analogous art, Zhu et al. disclose wherein the neural network includes a third neural subnetwork, the third neural subnetwork performing based on the extracted features as an output of the attention layer. (Zhu et al., Figs. 5A, 5B, 6A & 8, Pg. 8 ¶ 0096, Pg. 9 ¶ 0103 - 0105, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0128, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0159)
- With regards to claim 19, Huang et al. in view of Zhu et al. disclose the method according to claim 18, wherein the third neural subnetwork comprises applying of a softmax layer and obtaining the estimated probabilities as an output of the softmax layer, (Huang et al., Figs. 2, 5 & 7, Pg. 3 ¶ 0037 - 0040, Pg. 4 ¶ 0043, Pg. 5 ¶ 0062, Pg. 12 ¶ 0125, 0128 and 0133, Pg. 13 ¶ 0146 - 0150, Pg. 15 ¶ 0167) or wherein the third neural subnetwork performs the estimating of probabilities of information associated with the current node based on the context embedding related to the current node. (Huang et al., Figs. 2, 5 & 7, Pg. 4 ¶ 0040 - 0044, Pg. 5 ¶ 0057 - 0062, Pg. 12 ¶ 0128 - 0134, Pg. 13 ¶ 0146 - 0150, Pg. 15 ¶ 0163 - 0167)
- With regards to claim 20, Huang et al. in view of Zhu et al. disclose the method according to claim 1. ([See analysis of Claim 1 provided herein above.]) Huang et al. disclose a non-transitory computer-readable medium including code instructions, which, upon being executed by one or more processors, cause the one or more processors to execute (Huang et al., Abstract, Figs. 1, 6, 8 & 9, Pg. 1 ¶ 0005 - 0006, Pg. 7 ¶ 0079 - 0080 and 0085, Pg. 8 ¶ 0090 - 0091 and 0094, Pg. 15 ¶ 0169, Pg. 16 ¶ 0184 - 0185, Pg. 17 ¶ 0189 - 0194, Pg. 18 ¶ 0199 - 0201) the method according to claim 1. ([Huang et al. in view of Zhu et al. disclose the method according to claim 1, see analysis of claim 1 provided herein above.])
- With regards to claim 21, Huang et al. an apparatus for entropy encoding data of a three-dimensional point cloud, (Huang et al., Abstract, Figs. 1, 2 & 6 - 9, Pg. 1 ¶ 0005 - 0006 and 0018 - 0019, Pg. 2 ¶ 0023, Pg. 4 ¶ 0045 - 0050, Pg. 6 ¶ 0072 - 0075, Pg. 7 ¶ 0079 - 0080 and 0085 - 0087, Pg. 10 ¶ 0106 - 0108, Pg. 12 ¶ 0136, Pg. 16 ¶ 0184 - Pg. 17 ¶ 0188) comprising: a processing circuitry (Huang et al., Abstract, Figs. 1, 8 & 9, Pg. 1 ¶ 0005 - 0006, Pg. 7 ¶ 0079, Pg. 8 ¶ 0090 - 0091, Pg. 15 ¶ 0169, Pg. 16 ¶ 0184 - 0185, Pg. 17 ¶ 0189 - 0191 and 0193, Pg. 18 ¶ 0197 - 0201) configured to: for a current node in an N-ary tree-structure representing the three-dimensional point cloud: (Huang et al., Abstract, Figs. 2, 4 & 7, Pg. 1 ¶ 0005 and 0018 - 0019, Pg. 3 ¶ 0030, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0047, Pg. 4 ¶ 0050 - 0051, Pg. 10 ¶ 0109 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0135, Pg. 13 ¶ 0146 - Pg. 14 ¶ 0152) obtain a set of neighboring nodes of the current node; (Huang et a., Figs. 2, 4 & 7, Pg. 1 ¶ 0018 - 0019, Pg. 3 ¶ 0033 - 0036, Pg. 4 ¶ 0041 - 0044 and 0050 - 0051, Pg. 5 ¶ 0058 and 0062, Pg. 11 ¶ 0117 - 0123, Pg. 13 ¶ 0138 and 0142 - 0144, Pg. 15 ¶ 0163 - 0164 and 0167) extract features of the set of the neighboring nodes by applying a neural network including a layer; (Huang et al., Abstract, Figs. 2, 4, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0036 - Pg. 4 ¶ 0045, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0071 - 0073, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0162 - Pg. 15 ¶ 0167) estimate probabilities of information associated with the current node based on the extracted features; (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0037 - 0039, Pg. 4 ¶ 0042 - 0047 and 0050, Pg. 5 ¶ 0057 and 0062, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0155 - 0160, Pg. 15 ¶ 0163 - 0167) entropy encode the information associated with the current node based on the estimated probabilities. (Huang et al., Abstract, Figs. 2, 6 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 4 ¶ 0045 - 0047, Pg. 5 ¶ 0063, Pg. 6 ¶ 0074 - 0075, Pg. 11 ¶ 0114, Pg. 12 ¶ 0135, Pg. 13 ¶ 0150 - Pg. 14 ¶ 0152, Pg. 15 ¶ 0167 - 0168, Pg. 16 ¶ 0179 - 0180) Huang et al. fail to disclose explicitly a neural network including an attention layer. Pertaining to analogous art, Zhu et al. disclose an apparatus for entropy encoding data, (Zhu et al., Abstract, Figs.1, 3 & 8 - 10, Pg. 1 ¶ 0005 - 0007, Pg. 4 ¶ 0066, Pg. 5 ¶ 0073 - Pg. 6 ¶ 0077, Pg. 7 ¶ 0092 - Pg. 8 ¶ 0100, Pg. 9 ¶ 0104 - 0107, Pg. 16 ¶ 0160 - 0164) comprising: a processing circuitry (Zhu et al., Figs. 1 & 10, Pg. 1 ¶ 0005, Pg. 3 ¶ 0039, Pg. 5 ¶ 0074, Pg. 9 ¶ 0107 - Pg. 10 ¶ 0110, Pg. 16 ¶ 0161 - 0164, Pg. 19 ¶ 0186) configured to: for a current node in an N-ary tree-structure: (Zhu et al., Figs. 7A - 9, Pg. 7 ¶ 0087 - 0092, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0122 - Pg. 12 ¶ 0128, Pg. 15 ¶ 0154 - Pg. 16 ¶ 0160) extract features by applying a neural network including an attention layer; (Zhu et al., Figs. 5A, 5B, 6A & 7B- 9, Pg. 5 ¶ 0069 - 0073, Pg. 8 ¶ 0096, Pg. 9 ¶ 0102 - 0104, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0129, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0160) and entropy encode the information associated with the current node. (Zhu et al., Figs. 8 & 9, Pg. 5 ¶ 0069 - 0073, Pg. 7 ¶ 0087 - 0091, Pg. 8 ¶ 0094 - 0096, Pg. 9 ¶ 0102 - 0104, Pg. 12 ¶ 0129 - 0130, Pg. 16 ¶ 0155 - 0160) Huang et al. and Zhu et al. are combinable because they are both directed towards data encoding and decoding systems that utilize neural networks to entropy encode and decode input data. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Huang et al. with the teachings of Zhu et al. This modification would have been prompted in order to enhance the base device of Huang et al. with the well-known and applicable technique Zhu et al. applied to a comparable device. Utilizing a neural network including an attention layer to extract features, as taught by Zhu et al., would enhance the base device of Huang et al. by improving its ability to identify and extract the most significant and representative features from input data, the point cloud, thereby enhancing its ability to accurately and efficiently encode and decode input data. Furthermore, this modification would have been prompted by the teachings and suggestions of Huang et al. that their system can utilize multiple multi-layered perceptions during entropy encoding/decoding, that additional information can be used to improve encoding/decoding efficiency and that their system can additionally or alternatively utilize various other machine learning models, such as recurrent neural networks, convolutional neural networks or other forms of neural networks, see at least page 3 paragraphs 0036 - 0037, page 4 paragraph 0048, page 11 paragraph 0124 - page 12 paragraph 0125, page 17 paragraph 0195 and page 18 paragraph 0202 of Huang et al. Moreover, this modification would have been prompted by the teachings and suggestions of Zhu et al. that an improved rate-distortion loss was achieved by utilizing attention layers with a convolutional neural network (CNN) based coding system, see at least page 5 paragraphs 0069 and 0073 and page 12 paragraph 0129 of Zhu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that a neural network including an attention layer would be utilized by the base device of Huang et al. in order to improve its ability to identify and extract the most significant and representative features from input data, the point cloud, and thus enhance its ability to accurately and efficiently encode and decode input data. Therefore, it would have been obvious to combine Huang et al. with Zhu et al. to obtain the invention as specified in claim 21.
- With regards to claim 22, Huang et al. an apparatus for entropy decoding data of a three-dimensional point cloud, (Huang et al., Abstract, Figs. 1, 2 & 6 - 9, Pg. 1 ¶ 0005 - 0006 and 0018 - 0019, Pg. 2 ¶ 0024, Pg. 4 ¶ 0045 - 0050, Pg. 5 ¶ 0063, Pg. 6 ¶ 0074 - 0075, Pg. 7 ¶ 0079 - 0080 and 0085 - 0087, Pg. 10 ¶ 0106 - 0108, Pg. 12 ¶ 0136, Pg. 14 ¶ 0152, Pg. 16 ¶ 0181 - Pg. 17 ¶ 0188) comprising: a processing circuitry (Huang et al., Abstract, Figs. 1, 8 & 9, Pg. 1 ¶ 0005 - 0006, Pg. 7 ¶ 0079, Pg. 8 ¶ 0090 - 0091, Pg. 15 ¶ 0169, Pg. 16 ¶ 0184 - 0185, Pg. 17 ¶ 0189 - 0191 and 0193, Pg. 18 ¶ 0197 - 0201) configured to: for a current node in an N-ary tree-structure representing the three-dimensional point cloud: (Huang et al., Abstract, Figs. 2, 4 & 7, Pg. 1 ¶ 0005 and 0018 - 0019, Pg. 3 ¶ 0030, Pg. 3 ¶ 0033 - Pg. 4 ¶ 0047, Pg. 4 ¶ 0050 - 0051, Pg. 10 ¶ 0109 - Pg. 11 ¶ 0114, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0135, Pg. 13 ¶ 0146 - Pg. 14 ¶ 0152) obtain a set of neighboring nodes of the current node; (Huang et a., Figs. 2, 4 & 7, Pg. 1 ¶ 0018 - 0019, Pg. 3 ¶ 0033 - 0036, Pg. 4 ¶ 0041 - 0044 and 0050 - 0051, Pg. 5 ¶ 0058 and 0062, Pg. 11 ¶ 0117 - 0123, Pg. 13 ¶ 0138 and 0142 - 0144, Pg. 15 ¶ 0163 - 0164 and 0167) extract features of the set of the neighboring nodes by applying a neural network including a layer; (Huang et al., Abstract, Figs. 2, 4, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0036 - Pg. 4 ¶ 0045, Pg. 5 ¶ 0057 - 0062, Pg. 6 ¶ 0071 - 0073, Pg. 11 ¶ 0120 - Pg. 12 ¶ 0125, Pg. 12 ¶ 0130 - 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0162 - Pg. 15 ¶ 0167) estimate probabilities of information associated with the current node based on the extracted features; (Huang et al., Abstract, Figs. 2, 5 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 3 ¶ 0037 - 0039, Pg. 4 ¶ 0042 - 0047 and 0050, Pg. 5 ¶ 0057 and 0062, Pg. 6 ¶ 0071 - 0072, Pg. 11 ¶ 0124 - Pg. 12 ¶ 0134, Pg. 13 ¶ 0145 - 0150, Pg. 14 ¶ 0155 - 0160, Pg. 15 ¶ 0163 - 0167) entropy decode the information associated with the current node based on the estimated probabilities. (Huang et al., Figs. 2, 6 & 7, Pg. 1 ¶ 0005 and 0019, Pg. 2 ¶ 0024, Pg. 4 ¶ 0047, Pg. 5 ¶ 0057 and 0063, Pg. 6 ¶ 0074 - 0075, Pg. 11 ¶ 0114, Pg. 12 ¶ 0136, Pg. 14 ¶ 0152, Pg. 15 ¶ 0167 - 0168, Pg. 16 ¶ 0179 - 0183) Huang et al. fail to disclose explicitly a neural network including an attention layer. Pertaining to analogous art, Zhu et al. disclose an apparatus for entropy decoding data, (Zhu et al., Figs.1, 3 & 8 - 10, Pg. 1 ¶ 0005 - 0007, Pg. 4 ¶ 0066, Pg. 5 ¶ 0073 - Pg. 6 ¶ 0077, Pg. 7 ¶ 0092 - Pg. 8 ¶ 0100, Pg. 9 ¶ 0104 - 0107, Pg. 15 ¶ 0147 - 0151, Pg. 16 ¶ 0160 - 0164) comprising: a processing circuitry (Zhu et al., Figs. 1 & 10, Pg. 1 ¶ 0005, Pg. 3 ¶ 0039, Pg. 5 ¶ 0074, Pg. 9 ¶ 0107 - Pg. 10 ¶ 0110, Pg. 16 ¶ 0161 - 0164, Pg. 19 ¶ 0186) configured to: for a current node in an N-ary tree-structure: (Zhu et al., Figs. 7A - 9, Pg. 7 ¶ 0087 - 0092, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0122 - Pg. 12 ¶ 0128, Pg. 15 ¶ 0154 - Pg. 16 ¶ 0160) extract features by applying a neural network including an attention layer; (Zhu et al., Figs. 5A, 5B, 6A & 7B- 9, Pg. 5 ¶ 0069 - 0073, Pg. 8 ¶ 0096, Pg. 9 ¶ 0102 - 0104, Pg. 11 ¶ 0116 - 0119, Pg. 11 ¶ 0123 - Pg. 12 ¶ 0129, Pg. 13 ¶ 0135 - 0137, Pg. 16 ¶ 0155 - 0160) and entropy decode the information associated with the current node. (Zhu et al., Figs. 8 & 9, Pg. 5 ¶ 0069 - 0073, Pg. 7 ¶ 0087 - 0091, Pg. 8 ¶ 0094 - 0096, Pg. 9 ¶ 0102 - 0105, Pg. 11 ¶ 0120, Pg. 12 ¶ 0129 - 0130 and 0133, Pg. 15 ¶ 0147 - 0151, Pg. 16 ¶ 0158 - 0160) Huang et al. and Zhu et al. are combinable because they are both directed towards data encoding and decoding systems that utilize neural networks to entropy encode and decode input data. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Huang et al. with the teachings of Zhu et al. This modification would have been prompted in order to enhance the base device of Huang et al. with the well-known and applicable technique Zhu et al. applied to a comparable device. Utilizing a neural network including an attention layer to extract features, as taught by Zhu et al., would enhance the base device of Huang et al. by improving its ability to identify and extract the most significant and representative features from input data, the point cloud, thereby enhancing its ability to accurately and efficiently encode and decode input data. Furthermore, this modification would have been prompted by the teachings and suggestions of Huang et al. that their system can utilize multiple multi-layered perceptions during entropy encoding/decoding, that additional information can be used to improve encoding/decoding efficiency and that their system can additionally or alternatively utilize various other machine learning models, such as recurrent neural networks, convolutional neural networks or other forms of neural networks, see at least page 3 paragraphs 0036 - 0037, page 4 paragraph 0048, page 11 paragraph 0124 - page 12 paragraph 0125, page 17 paragraph 0195 and page 18 paragraph 0202 of Huang et al. Moreover, this modification would have been prompted by the teachings and suggestions of Zhu et al. that an improved rate-distortion loss was achieved by utilizing attention layers with a convolutional neural network (CNN) based coding system, see at least page 5 paragraphs 0069 and 0073 and page 12 paragraph 0129 of Zhu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that a neural network including an attention layer would be utilized by the base device of Huang et al. in order to improve its ability to identify and extract the most significant and representative features from input data, the point cloud, and thus enhance its ability to accurately and efficiently encode and decode input data. Therefore, it would have been obvious to combine Huang et al. with Zhu et al. to obtain the invention as specified in claim 22.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Cricrì et al. U.S. Publication No. 2024/0289590 A1; which is directed towards an image and video compression method and system, wherein a neural network with attention layers is utilized to entropy encode image and/or video data.
Nalci et al. U.S. Publication No. 2023/0096567 A1; which is directed towards neural-network-based image and video coding techniques, wherein a neural network with integrated attention layers is utilized to entropy encode image and/or video data.
Urtasun et al. U.S. Publication No. 2020/0160151 A1; which is directed towards systems and methods for compressing two-dimensional and/or three-dimensional image data, wherein a combination of a machine-learned feature extraction model, a machine-learned attention model and a machine-learned encoding model is utilized to perform compression operations on two-dimensional and/or three-dimensional image data.
Yuan et al. U.S. Publication No. 2024/0177353 A1; which is directed towards a point cloud processing method, wherein a multihead autocorrelation attention mechanism is utilized in a point cloud up-sampling method.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ERIC RUSH whose telephone number is (571) 270-3017. The examiner can normally be reached 9am - 5pm Monday - Friday.
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, Andrew Bee can be reached at (571) 270 - 5183. 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.
/ERIC RUSH/Primary Examiner, Art Unit 2677