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 Arguments
Applicant’s arguments, filed 11/24/2025, have been fully considered and the examiner’s responses are given below.
The 35 U.S.C. 112(b) rejections are partially withdrawn, and new rejections are presented below.
The 35 U.S.C. 101 rejections are withdrawn.
Applicant has claimed a practical application of using the compressed HD map to perform autonomous navigation.
The 35 U.S.C. 103 rejections are withdrawn, however new grounds are presented below.
Applicant’s amendments to the independent claims alter the scope of the claims, therefore new prior art has been applied and applicant’s arguments are moot.
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 1-20 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.
Regarding claims 1, 14, and 18, “the compressed HD map” lacks antecedent basis, therefore this claim is indefinite. For the purposes of examination, Examiner has interpreted “the compressed HD map” to mean “the compressed graph HD map”.
Regarding claims 2-13, these claims depend from claim 1 and are therefore rejected for the same reason as claim 1 above, as they do not cure the deficiencies of claim 1 noted above.
Regarding claims 15-17, these claims depend from claim 14 and are therefore rejected for the same reason as claim 14 above, as they do not cure the deficiencies of claim 14 noted above.
Regarding claims 19-20, these claims depend from claim 18 and are therefore rejected for the same reason as claim 18 above, as they do not cure the deficiencies of claim 18 noted above.
Regarding claim 2, 15, and 19, “the compressed graph multi-edge tree” lacks antecedent basis, therefore these claims are indefinite. For the purposes of examination, Examiner has interpreted “the compressed graph multi-edge tree” to mean “a compressed graph multi-edge tree”.
Regarding claim 3, this claim depends from claim 2 and is therefore rejected for the same reason as claim 2 above, as it does not cure the deficiencies of claim 2 noted above.
Regarding claims 19-20, these claims recite “the media of claim 18”. It is unclear if “the media” refers to the “computer-readable non-transitory storage media” of claim 18 or refers to other media in claim 18, therefore this claim is indefinite. For the purposes of examination, Examiner has interpreted “the media of claim 18” in claims 19-20 to mean “the computer-readable non-transitory storage media of claim 18”.
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 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.
Claims 1, 5, 11-14, 16-18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Kum (US 20250139986 A1, cited in a previous office action) in view of Wu (US 20250200283 A1), Banerjee (US 20240402725 A1), Ehsanibenafati (US 20220144304 A1, cited in a previous office action), and Cserna (US 20220290997 A1, cited in a previous office action).
Regarding claim 1, Kum discloses a method comprising, by an electronic device (Paragraph 0009, 0037, 0093-0094);
accessing sensor data captured by one or more mobile devices operating in a plurality of regions in an environment (Paragraph 0069-0070, 0089-0092);
generating, for each of the plurality regions, a plurality of graph embeddings associated with each of the mobile devices (Paragraphs 0068, 0075-0076, 0089-0092);
based on a compressed graph associated with each of the mobile devices constructed from the sensor data (Paragraphs 0068, 0075-0076, 0089-0092; compressing is done by fusing vertex coordinates and edge information);
the compressed graph comprises a plurality of nodes and a plurality of compressed edges connecting the nodes across a plurality of layers (Paragraph 0077; Layers is mapped to multi-layer perceptron (MLP));
generating, for each of the plurality of regions based on the plurality of graph embeddings associated with each of the mobile devices, a refined graph associated with each of the mobile devices (Paragraphs 0075-0079, 0089-0092);
by reconfiguring one or more edges of the plurality of compressed edges across the plurality of layers in the compressed graph associated with each of the mobile devices (Paragraphs 0076-0079, 0089-0092);
associated with each of the mobile devices (Paragraph 0069-0070, 0089-0092);
generating a graph high-definition (HD) map (Paragraph 0058, 0091);
and using the compressed HD map to perform autonomous navigation and control (Paragraph 0091-0092).
Kum does not specifically state quantizing text tokens of text information to select compressible tokens; wherein the text information is detected from the environment based on the sensor data.
However, Wu teaches quantizing text tokens of text information to select compressible tokens (Paragraphs 0090, 0096, 0127-0131; tokenized description… adaptive representations of spatial locations can be used to reduce the amount of precision needed and allow for fewer and/or smaller tokens);
wherein the text information is detected from the environment based on the sensor data (Paragraphs 0090, 0094-0096, 0116; “an HD map (which may be represented using an occupancy map generated from any type of sensor data, such as image data, LiDAR data, RADAR data, etc., in embodiments) may have various layers—such as planning layers, semantic data layers, sensor-specific layers (e.g., RADAR layers, LiDAR layers, camera layers, etc.), and/or other layer types. To convert the map representation to a language-based representation, such as a sequential, tokenized text string, one or more of these layers may be used”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with detecting text information from the environment and quantizing text tokens of Wu with a reasonable expectation of success. One of ordinary skill in the art would understand that objects in an environment can be represented using text tokens. Quantizing the text tokens allows for lightweight and compact descriptions for faster processing. One would have been motivated to combine Kum with Wu as this achieves lightweight and efficient recognition of the environment. As stated in Wu, “The use of a tokenized description to store this information also helps to retain this additional information as the description is lightweight and compact, and can be sparser and more discrete, relative to other environment representations. In at least one embodiment, however, the information to be stored in such a representation can be limited to only information that is useful, or even necessary, for a task to be performed, such as autonomous navigation. By excluding information that is not needed for such a task, the tokenized description can be even more compact and allow for faster processing” (Paragraph 0090).
Kum does not specifically state identifying and removing non-prevalent edges from the plurality of compressed edges associated with each refined graph based on the selected compressible tokens.
However, Banerjee teaches identifying and removing non-prevalent edges from the plurality of compressed edges associated with each refined graph based on the selected compressible tokens (Paragraphs 0041-0043; “node and edge duplicates are avoided, and low weight and noisy edges are removed”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with removing non-prevalent edges associated with each refined graph based on the compressible tokens of Banerjee with a reasonable expectation of success. One of ordinary skill in the art would understand that when a robot creates a global graph from refined graphs, edges that are low weight (only seen in 1 refined graph) can be removed. High weight edges seen in more than 1 graph are kept in the global graph because these are more important. One would have been motivated to combine Kum with Banerjee as this achieves efficient navigation. As stated in Banerjee, “The semantic map of the context is used in navigation for efficiency” (Paragraph 0044).
Kum does not specifically state associated with each of the mobile devices based on a fusion of the one or more refined graphs associated with the plurality of regions for each of the mobile devices.
However, Ehsanibenafati teaches associated with each of the mobile devices based on a fusion of the one or more refined graphs associated with the plurality of regions for each of the mobile devices (Abstract, Paragraphs 0021-0022, 0052).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with fusing a plurality of refined graphs of Ehsanibenafati with a reasonable expectation of success. One of ordinary skill in the art would understand that it is possible to use graph maps from a variety of sources to improve graph accuracy. The fusion of multiple sources of graph data allows errors from individual sources to be reduced. Also comparing the fused graph to individual graphs allows for confidence and reliability metrics to be determined, improving navigation. One would have been motivated to combine Kum with Ehsanibenafati as this achieves improved graph HD map accuracy. As stated in Ehsanibenafati, “to improve the accuracy of the lane graph function, the lane graph fusion module 274 can utilize different sources of lane graph data, such as HD map-based lane graph and perception-based lane graph data, that can be fused together” (Paragraphs 0054-0055).
Kum does not specifically state identifying, based on the graph HD map associated with each of the mobile devices, a plurality of prominent nodes and a plurality of prominent edges connecting the prominent nodes based on trace activations associated with the prominent edges; generating a compressed graph HD map for the environment based on the prominent nodes and prominent edges associated with the one or more mobile devices and environmental information associated with the environment.
However, Cserna teaches identifying, based on the graph HD map associated with each of the mobile devices, a plurality of prominent nodes and a plurality of prominent edges connecting the prominent nodes based on trace activations associated with the prominent edges (Paragraph 0128-0134; Prominence is based on average distance from drive log);
generating a compressed graph HD map for the environment based on the prominent nodes and prominent edges associated with the one or more mobile devices and environmental information associated with the environment (Paragraph 0128-0134).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with identifying prominent nodes and prominent edges, and compressing the graph HD map based on these prominent nodes and edges of Cserna with a reasonable expectation of success. One of ordinary skill in the art would understand that prominent nodes and edges can be identified based on the distance to the vehicle and its trajectories. Less important nodes and edges are usually located further away from where the vehicle and its trajectories are located. These less important nodes and edges can be removed from the graph HD map in order to reduce computational load. One would have been motivated to combine Kum with Cserna as compressing a graph reduces the computational size in subsequent processes. As stated in Cserna, “the graph 1312 has been pruned or reduced to include edges that are determined to be relevant to the vehicle or otherwise useful. By reducing the edges in a graph to those edges determined to be relevant, useful, or otherwise important to path planning, the subsequent computations performed by a planning module are also reduced. Trajectories derived from a reduced graph that includes edges labeled as useful exhibit the same high quality as trajectories derived from a dense graph” (Paragraph 0117).
Regarding claim 5, Kum discloses associated with each of the mobile devices (Paragraphs 0058, 0068-0069, 0075-0076, 0089-0092).
Kum does not specifically state calculating, based on one or more of an active learning model or an out-of-distribution model, a plurality of trace activations for the plurality of graph embeddings based on one or more of a distance metric or a quantitative similarity measure; generating the refined graph associated with each of the mobile devices for each of the plurality of regions is further based on the plurality of trace activations.
However, Cserna teaches calculating, based on one or more of an active learning model or an out-of-distribution model, a plurality of trace activations for the plurality of graph embeddings based on one or more of a distance metric or a quantitative similarity measure (Paragraph 0128-0134);
generating the refined graph associated with each of the mobile devices for each of the plurality of regions is further based on the plurality of trace activations (Paragraph 0128-0134).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with calculating trace activations and generating refined graphs based on the trace activations of Cserna with a reasonable expectation of success. One of ordinary skill in the art would understand that trace activations can based on the distance from the vehicle. The further the embedding is from the vehicle, the lower the trace activation and the less important it is. These nodes and edges can be removed from the graph HD map because they are less relevant. One would have been motivated to combine Kum with Cserna as compressing a graph reduces the computational size in subsequent processes. As stated in Cserna, “the graph 1312 has been pruned or reduced to include edges that are determined to be relevant to the vehicle or otherwise useful. By reducing the edges in a graph to those edges determined to be relevant, useful, or otherwise important to path planning, the subsequent computations performed by a planning module are also reduced. Trajectories derived from a reduced graph that includes edges labeled as useful exhibit the same high quality as trajectories derived from a dense graph” (Paragraph 0117).
Regarding claim 11, Kum discloses the one or more mobile devices are one or more vehicles, and wherein the method further comprises (Paragraph 0069-0070, 0089-0092);
determining static environmental information associated with the environment, wherein the static environmental information comprises one or more of lane information or road-level information (Paragraph 0048);
determining, based on the sensor data, dynamic environmental information associated with the environment, wherein the dynamic environmental information comprises one or more of environmental localization, scene understanding, textual inference, or road obstruction (Paragraph 0048, 0051);
the environmental information comprises the static environmental information and the dynamic environmental information (Paragraph 0069-0073, 0091).
Regarding claim 12, Kum discloses the one or more mobile devices are one or more vehicles, and wherein the method further comprises (Paragraph 0069-0070, 0089-0092);
executing one or more processing tasks based on the compressed graph HD map, wherein the one or more processing tasks comprise one or more of navigation, steering, acceleration, or deceleration (Paragraph 0091-0092).
Regarding claim 13, Kum discloses the one or more mobile devices are one or more vehicles (Paragraph 0069-0070, 0089-0092);
the sensor data comprises one or more of LiDAR data, image data, GPS data, inertial-measurement-unit (IMU) data, or radar data (Paragraph 0057, 0089).
Regarding claim 14, Kum discloses one or more non-transitory computer-readable storage media including instructions (Paragraphs 0039-0041);
one or more processors coupled to the storage media, the one or more processors configured to execute the instructions to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 1. The electronic device taught/disclosed in claim 14 can be clearly performed with the method of claim 1. Therefore, claim 14 is rejected under the same rationale.
Regarding claim 16, Kum discloses the one or more processors are further configured to execute the instructions to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 5. The electronic device taught/disclosed in claim 16 can be clearly performed with the method of claim 5. Therefore, claim 16 is rejected under the same rationale.
Regarding claim 17, Kum discloses the one or more processors are further configured to execute the instructions to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 11. The electronic device taught/disclosed in claim 17 can be clearly performed with the method of claim 11. Therefore, claim 17 is rejected under the same rationale.
Regarding claim 18, Kum discloses a computer-readable non-transitory storage media comprising instructions executable by a processor to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 1. The computer-readable non-transitory storage media taught/disclosed in claim 18 can be clearly performed with the method of claim 1. Therefore, claim 18 is rejected under the same rationale.
Regarding claim 20, Kum discloses the media further comprises instructions executable by the processor to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 11. The computer-readable non-transitory storage media taught/disclosed in claim 20 can be clearly performed with the method of claim 11. Therefore, claim 20 is rejected under the same rationale.
Claims 2-3, 15, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Kum, Wu, Banerjee, Ehsanibenafati, and Cserna, as applied to claims 1, 14, and 18 above, and further in view of Patodia (US 20220198283 A1, cited in a previous office action).
Regarding claim 2, Kum discloses generating the compressed graph associated with each of the mobile devices, wherein the generation comprises (Paragraphs 0068, 0075-0076, 0089-0092);
generating, for each of the plurality regions based on the sensor data (Paragraph 0070, 0089-0092);
the plurality of nodes correspond to a plurality of objects in the region (Paragraphs 0051-0052);
one or more of the nodes are connected by one or more of the edges in each of the layers (Paragraphs 0076-0079).
Kum does not specifically state a graph multi-edge tree; comprising the plurality of nodes and a plurality of edges connecting the nodes across the plurality of layers; the one or more edges in each of the layers represent a respective level of relationship between the one or more nodes in that layer; generating, for each of the plurality of regions, the compressed graph multi-edge tree from the corresponding graph multi-edge tree by extracting a plurality of high-activating edges across the plurality of layers from the graph multi-edge tree, wherein the compressed graph multi-edge tree comprises the plurality of compressed edges comprising the plurality of high-activating edges across the plurality of layers.
However, Patodia teaches a graph multi-edge tree (Paragraphs 0046, Figs. 2A-2B);
the plurality of nodes and a plurality of edges connecting the nodes across the plurality of layers (Paragraphs 0049-0051, Figs. 2A-2B);
the one or more edges in each of the layers represent a respective level of relationship between the one or more nodes in that layer (Paragraphs 0049-0051, Figs. 2A-2B);
generating, for each of the plurality of regions, the compressed graph multi-edge tree from the corresponding graph multi-edge tree by extracting a plurality of high-activating edges across the plurality of layers from the graph multi-edge tree, wherein the compressed graph multi-edge tree comprises the plurality of compressed edges comprising the plurality of high-activating edges across the plurality of layers (Paragraph 0046; High-activation is based on most frequently used tree structures).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with generating a graph multi-edge tree and compressing it based on high-activating edges of Patodia with a reasonable expectation of success. One of ordinary skill in the art would understand that a multi-edge decision tree can be setup to aid in vehicle navigation. Compressing the tree reduces the amount of nodes and edges which speeds up processing speeds. One would have been motivated to combine Kum with Patodia as this enhances vehicle navigation. As stated in Patodia, “the decision tree server 104 may determine which of the decision tree structures 142 are most frequently used for the policy and retain all or some (e.g., 5, 10, 50, etc.) of those structures 142 in the repository 110, which may reduce the time required to provide those decision tree structures 142 to the requesting application 122” (Paragraph 0046).
Regarding claim 3, Kum discloses generating the compressed graph.
Kum does not specifically state the level of relationship comprises one or more of a primary relationship, a secondary relationship, or a tertiary relationship.
However, Patodia teaches the level of relationship comprises one or more of a primary relationship, a secondary relationship, or a tertiary relationship (Paragraphs 0049-0051, Figs. 2A-2B; Primary, secondary, or tertiary is mapped to edges between L1 and L2, L2 and L3, L3 and L4).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with a primary, secondary, and tertiary level of relationship of Patodia with a reasonable expectation of success. One of ordinary skill in the art would understand that given multiple conditions that depend on each other, a graph multi-edge tree would require a hierarchy of levels. This allows decision making in navigation to be fully represented. One would have been motivated to combine Kum with Patodia as this achieves generating and using a decision tree structure. As stated in Patodia, “for a rule with multiple conditions, a corresponding decision tree structure 142 may include multiple levels, one for each of the conditions” (Paragraph 0049).
Regarding claim 15, Kum discloses the one or more processors are further configured to execute the instructions to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 2. The electronic device taught/disclosed in claim 15 can be clearly performed with the method of claim 2. Therefore, claim 15 is rejected under the same rationale.
Regarding claim 19, Kum discloses instructions executable by the processor to (Paragraphs 0039-0041).
All other limitations have been examined with respect to the method in claim 2. The computer-readable non-transitory storage media taught/disclosed in claim 19 can be clearly performed with the method of claim 2. Therefore, claim 19 is rejected under the same rationale.
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Kum, Wu, Banerjee, Ehsanibenafati, and Cserna, as applied to claim 1 above, and further in view of Joshi (US 20250128734 A1, cited in a previous office action).
Regarding claim 4, Kum discloses each of the plurality of graph embeddings associated with each of the mobile devices comprises one or more of edge-level vector information or node-level vector information (Paragraphs 0075-0079);
each of the plurality of graph embeddings is reconfigurable (Paragraphs 0075-0079).
Kum does not specifically state each of the plurality of graph embeddings is associated with a dynamic vector length.
However, Joshi teaches each of the plurality of graph embeddings is associated with a dynamic vector length (Paragraph 0016).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with graph embeddings associated with a dynamic vector length of Joshi with a reasonable expectation of success. One of ordinary skill in the art would understand that using dynamic vector lengths allows the environment to be represented. This model representation may also be transferrable to other computers or machine learned models to train them and help them understand the environment. One would have been motivated to combine Kum with Joshi as this achieves increased modeling efficiency. As stated in Joshi, “encoding in meaningful latent space (e.g., an arbitrary-length tensor or vector) that represents the machine learned model's understanding of the environment and the relevance of features therein to one another. The representation may be transferrable to multiple online or offline computer systems or machine learned models associated with the vehicle in place of separate respective processes to train models and then interpret and process sensor data to obtain an understanding of the environment. In this way, there is reduced redundancy of work and reduced overhead” (Paragraph 0016).
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Kum, Wu, Banerjee, Ehsanibenafati, and Cserna, as applied to claim 1 above, and further in view of Kompella (US 20220179857 A1, cited in a previous office action).
Regarding claim 6, Kum discloses associated with each of the mobile devices (Paragraphs 0058, 0068-0069, 0075-0076, 0089-0092).
Kum does not specifically state detecting the text information from the environment; generating, based on the text information, a plurality of text tokens; determining a plurality of text-graph correspondences indicating a plurality of mappings between one or more of the graph embeddings and one or more of the text tokens.
However, Kompella teaches detecting the text information from the environment (Paragraphs 0051-0054, Fig. 5);
generating, based on the text information, a plurality of text tokens (Paragraphs 0051-0054, Fig. 5);
determining a plurality of text-graph correspondences indicating a plurality of mappings between one or more of the graph embeddings and one or more of the text tokens (Paragraphs 0051-0054, Fig. 5).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify the invention of Kum with detecting the text information of the environment, generating text tokens, and a determining mappings between graph embeddings and text tokens of Kompella with a reasonable expectation of success. One of ordinary skill in the art would understand that vehicles may need to recognize their environment in order to navigate properly. Incorporating text into graph data improves the vehicle’s machine learning model and recognition of the environment. One would have been motivated to combine Kum with Kompella as this achieves an increased understanding of the environment. As stated in Kompella, “Machine learning, for instance, enables service providers to extract underlying spatial and/or semantic relationships between locations and to classify or make predictions based on those relationships or underlying structure. In many cases, the locations can also be associated with unstructured multi-modal data such as text, images, etc. As a result, service providers face significant technical challenges to encode and represent locations along with their semantic/structural relationships and related multi-modal data in a way that can be used for machine learning tasks” (Paragraph 0001).
Allowable Subject Matter
Claims 7-10 would be allowable if rewritten to overcome the rejection(s) under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action and to include all of the limitations of the base claim and any intervening claims.
The following is a statement of reasons for the indication of allowable subject matter:
Dependent claim 7 recites (emphasis added):
“The method of Claim 6, further comprising:
determining a token quantization to identify one or more of the plurality of text tokens to compress based on a fusion process of the plurality of text tokens and the plurality of graph embeddings associated with each of the mobile devices, wherein the compressible tokens comprise the identified text tokens to compress.”
The prior art does not teach, disclose, or otherwise render obvious the above-noted features of the claims.
Kum (US 20250139986 A1) discloses graph embeddings associated with each of the mobile devices.
Kum, however, does not specifically state determining a token quantization to identify one or more of the plurality of text tokens to compress based on a fusion process of the plurality of text tokens and the plurality of graph embeddings.
Garg (US 10885452 B1) teaches identifying, based on the identified text tokens to compress, one or more non-prevalent edges of the plurality of compressed edges; and removing the one or more non-prevalent edges from the plurality of compressed edges (Col. 2 Line 38 – Col. 3 Line 13, Col. 12 Line 4 – Col. 12 Line 27, Claim 1).
Garg, however, does not specifically state determining a token quantization to identify one or more of the plurality of text tokens to compress based on a fusion process of the plurality of text tokens and the plurality of graph embeddings.
These differences between the subject matter of claim 7 and the prior art are not taught or otherwise rendered obvious by any available evidence in the remaining prior art. Accordingly, claim 7 recites allowable subject matter.
The following is a statement of reasons for the indication of allowable subject matter:
Dependent claim 8 recites:
“The method of Claim 6, further comprising:
determining, for each of the plurality regions, a joint attention associated with each of the mobile devices based on the plurality of text-graph correspondences.”
The prior art does not teach, disclose, or otherwise render obvious the above-noted features of the claims.
Kum (US 20250139986 A1) discloses a text-graph for mobile devices in each of the plurality of regions.
Kum, however, does not teach a joint attention associated with each of the mobile devices.
Huang (CN 115661594 A) teaches a cross attention based on the plurality of text-graph correspondences.
Huang, however, does not teach a joint attention based on the plurality of text-graph correspondences.
These differences between the subject matter of claim 8 and the prior art are not taught or otherwise rendered obvious by any available evidence in the remaining prior art. Accordingly, claim 8 recites allowable subject matter.
Claims 9-10 recite allowable subject matter based upon their dependency from claim 8.
Any comments considered necessary by applicant must be submitted no later than the payment of the issue fee and, to avoid processing delays, should preferably accompany the issue fee. Such submissions should be clearly labeled “Comments on Statement of Reasons for Allowance.”
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner
should be directed to Matthew Ho whose telephone number is (571) 272-1388. The examiner can
normally be reached on Mon-Thurs 9:00-5:30 EST.
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, Navid Z Mehdizadeh can be reached on (571)-272-7691. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications are available through Private PAIR only. For more information about the PAIR system, see https://ppairmy.uspto.gov/pair/PrivatePair. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at (866) 217-9197 (tollfree). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call (800) 786-9199 (IN USA OR CANADA) or (571) 272-1000.
/MATTHEW HO/ Examiner, Art Unit 3669
/NAVID Z. MEHDIZADEH/Supervisory Patent Examiner, Art Unit 3669