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 .
Examiner Notes that the fundamentals of the rejections are based on the broadest reasonable interpretation of the claim language. Applicant is kindly invited to consider the reference as a whole. References are to be interpreted as by one of ordinary skill in the art rather than as by a novice. See MPEP 2141. Therefore, the relevant inquiry when interpreting a reference is not what the reference expressly discloses on its face but what the reference would teach or suggest to one of ordinary skill in the art.
Status of the Claims
This Final Action is in response to Applicant’s amendment of 04 May 2026. Claims 1, 3-9, 11-17 & 19-20 are pending and have been considered as follows. Claims 2, 10 and 18 have been cancelled.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 05/04/2026 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to Arguments
Applicant’s amendments and/or arguments with respect to the objection of the drawings as set forth in the office action of 02 March 2026 have been considered and are NOT persuasive. Please see the maintained objection for further elaboration by the examiner.
Applicant’s amendments and/or arguments with respect to the rejection of Claims 1-7, 9-15 and 17-20 under 35 USC 101 as set forth in the office action of 02 March 2026 have been considered and are NOT persuasive. Specifically, Applicant argues:
Claims 1-7, 9-15 and 17-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Specifically, the Examiner asserts that the claims constitute a mental process. Applicant respectfully disagrees.
First, Applicant submits that, as the Supreme Court has cautioned, "describing the claims at such a high level of abstraction and untethered from the language of the claims all but ensures that the exceptions to § 101 swallow the rule." Alice Corp., 134 S. Ct. at 2354 (citing Mayo Collaborative Servs. v. Prometheus Labs. Inc., 566 US 66, 71, 101 USPQ2d 1961, 1965 (2012)). This is because, at some level, all inventions "embody, use, reflect, rest upon, or apply laws of nature, natural phenomena, or abstract ideas." Id. Thus, simply reciting that an abstract idea is allegedly present or recited by limitations in a claim is insufficient to establish that the claim as a whole is directed to an abstract idea under the test set forth in Alice Corp. (See also MPEP j 2106.04()(1), citing Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1335, 118 USPQ2d 1684, 1688 (Fed. Cir. 2016) ("The 'directed to' inquiry, therefore, cannot simply ask whether the claims involve a patent-ineligible concept, because essentially every routinely patent-eligible claim involving physical products and actions involves a law of nature and/or natural phenomenon")). Examiners should accordingly be careful to distinguish claims that recite an exception (which require further eligibility analysis) and claims that merely involve an exception (which are eligible and do not require further eligibility analysis). Here, Applicant's claims at most merely involve an exception and can only be alleged to recite an abstract idea when considered at too high a level of abstraction; therefore, Applicant's claims are in fact not directed to a judicial exception. For example, claim 1 recites, inter alia, executing various steps using a processor.
Second, Applicant respectfully submits that the mere possibility that a limitation can be performed as a mental step (i.e., by a human) is insufficient to establish that a claim as a whole is directed to an abstract idea. Applicant respectfully submits that using a process to execute various processing steps in the manner recited, when read in combination with other features of the claims, is not a judicial exception, but instead corresponds to non-excepted subject matter that is not and cannot be practically performed in the human mind and thus is not a "mental process." MPEP 2106.04(a)(2)(III). This is especially true in the context of performing complex computations and processing. As clarified in a recent memo ("Reminders on evaluating subject matter eligibility of claims under 35 U.S.C. § 101," United States Patent and Trademark Office (August 4, 2025)): The mental process grouping is not without limits. Examiners are reminded not to expand this grouping in a manner that encompasses claim limitations that cannot practically be performed in the human mind. The MPEP and the AI-SME Update provide examples of claim limitations that cannot be practically performed in the human mind. Claim limitations that encompass AI in a way that cannot be practically performed in the human mind do not fall within this grouping. At least one limitation in claim 1 as amended requires the use of a neural network and cannot be practically performed in the human mind. For at least this reason, Applicant submits that the claims do not recite a judicial exception and, as a whole, are eligible without any further eligibility analysis.
Third, even if, arguendo, the claims can be considered to recite a judicial exception, these claims integrate the alleged judicial exception into a practical application. "[A] claim that recites a judicial exception is not directed to that judicial exception, if the claim as a whole integrates the recited judicial exception into a practical application of that exception." MPEP j 2106.04(II)(A)(2). This is because integrating the judicial exception into a practical application will "apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the judicial exception." MPEP § 2106.04(d).
MPEP 2106.04(d)(I) identifies "a number of considerations as relevant to the evaluation of whether the claimed additional elements demonstrate that a claim is directed to patent-eligible subject matter." The analysis involves "evaluating a set of judicial considerations to determine if the claim is eligible." MPEP 2106.04(I). For example, one consideration is whether the claim provides an improvement to a technology or technical field. Here, using a processor to perform steps to generating an HD map and transmit the HD map to a vehicle for use in controlling autonomous functions in the manner recited imposes meaningful limits on the claim. Accordingly, the claim recites specific improvements to method for using a processor and neural network to generate an HD map used for controlling autonomous functions of a vehicle. Thus, Applicant's claims recite additional elements beyond any judicial exception that, at least in combination, integrate the alleged exception into a practical application because the additional elements reflect, for example, "an improvement in the functioning of a computer, or an improvement to other technology or technical field." (Emphasis added). See MPEP j 2106.04(d)(I). For example, the claims provide specific improvements over prior systems and recite additional elements that "apply[] or use[] the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment. Id. Thus, even where Applicant's claims allegedly recite a judicial exception, these claims clearly are not directed to the judicial exception and instead are directed to eligible subject matter.
In view of the foregoing, Applicant respectfully submits that claims 1, 9, and 17 are directed to statutory subject matter. Claims 3-8, 11-16, 19, and 20 ultimately depend from claims 1, 9, and 17 and likewise are directed to statutory subject matter.
The Examiners Response:
Examiner has carefully considered Applicant’s amendments and arguments and respectfully disagrees. Regarding the claimed invention, the claims merely recite obtaining information such as SD map data and aerial images, using the received information to predict and label features in the received information, generating an HD map, transmitting the map to a vehicle and receiving additional information to generate an additional HD map. The claims do have mental processes that can be performed in the human mind or with pen and paper as the current claims merely involve receiving information, predicting features and generating map data, furthermore, the inclusion of a computer/processor does not integrate the abstract idea into a patent eligible invention, See Alice Corp. Pty. Ltd. v. CLS Bank Int'!, 573 U.S. at 223 ("[T]he mere recitation of a generic computer cannot transform a patent-ineligible abstract idea into a patent-eligible invention."). The device(s) and processor(s) are recited at a high level of generality and merely automates the steps. Receiving data and transmitting data is insignificant extra-solution activity or well-understood routine activities.
Furthermore, the addition of the “neural network” as an additional element is not sufficient to claim a practical application or amount to significantly more than a judicial exception as the limitation represents no more than mere instructions to apply the judicial exception on a computer and it can also be viewed as nothing more than an attempt to generally link the use of the judicial exception to the technological environment of computers.
Finally, it is not clear what the improvement would be for using a processor and neural network to generate an HD map used for controlling autonomous functions of a vehicle and for the improvement to be indicative of integration into a practical application, it cannot be an improvement to an abstract idea, it must be "to the functioning of a computer, or to any other technology or technical field (MPEP 2106.05(a)). As such, even in combination, these additional elements, under broadest reasonable interpretation, do not integrate the abstract idea into practical application because they do not impose any meaningful limitations on practicing the abstract idea.
Applicant’s amendments and/or arguments with respect to the rejection of claims 1-20 under 35 USC 103 have been considered and are NOT persuasive. Specifically, Applicant argues:
Claims 1-3, 9-11 and 17-18 are rejected under 35 U.S.C. § 103 as being unpatentable over U.S. Publication No. 2025/0012600 to Buslaev, hereinafter Buslaev in view of U.S. Publication No. 2023/0349716 to Wang, hereinafter Wang and in further view of U.S. Publication No. 2019/0147331 to Arditi, hereinafter Arditi. Claims 4-5, 12-13 and 19 are rejected under 35 U.S.C. § 103 as being unpatentable over Buslaev in view of Wang, Arditi and in further view of Non- Patented Literature (NPL) Li "Neural Scene Flow Prior", 2021, hereinafter Li. Claims 6, 14 and 20 are rejected under 35 U.S.C. § 103 as being unpatentable over Buslaev, Wang, Arditi, Li and in further view of NPL Liu "VectorMapNet: End-to-end Vectorized HD Map Learning", 2023, hereinafter Liu. Claims 7-8 and 15-16 are rejected under 35 U.S.C. § 103 as being unpatentable over Buslaev, Wang, Arditi, Li, Liu and in further view of U.S. Publication No. 2022/0306156 to Wray, hereinafter Wray. These rejections are respectfully traversed.
Claim 1 recites, inter alia, in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
Buslaev, either singly or in combination with any other reference, fails to disclose generating a second HD map in the manner recited in claim 1.
The Examiner relies on Buslaev to generally disclose generating a first HD map based on an SD map and one or more aerial images (Office Action, Pages 9-10). Further, in rejecting claim 6, the Examiner suggests that Paragraphs [0059], [0060], and [0072] of Buslaev disclose "generating a second/updated HD map at the vehicle with new sensor information." (Office Action, Pages 17-18). However, Applicant respectfully submits that the relied upon portions of Buslaev do not disclose:
in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
For example, Paragraph [0059] of Buslaev only generally discloses that "different types of external data and/or client data may be utilized" for identifying map features, but fails to disclose, specifically, generating, in real-time, a second HD map based on the received images, the one or more NSPs, and the first HD map. Similarly, Paragraph [0060] only generally refers to an example workflow for generating an HD map database. Applicant notes that Paragraph [0062] of Buslaev at best only generally discloses that maps "having the SD map layers" can be augmented to generate HD map layers. In other words, these cited portions simply disclose the generation of an HD map, not the generation of a second HD map, in real-time while navigating the environment with the vehicle, based on the received images, NSPs, and the first HD map as required by claim 1.
Therefore, claim 1 is allowable for at least these reasons.
The Examiner’s Response:
Examiner has carefully considered Applicant's arguments and respectfully disagrees. Examiner agrees that Buslaev may not explicitly teach the limitation of “in real-time….”, however Examiner believes that the previously cited art of Arditi does teach the limitation after further consideration. Regarding Arditi teaching “in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map”, examiner respectfully disagrees that Arditi does not teach the limitation. Specifically, Arditi teaches obtaining real-time data from a vehicle navigating an environment and using the real-time data, CNN/DCNN and a previously existing HD map to generate a second/updated HD map, see [044] “Particular embodiments described herein relates to how an HD map may be updated periodically in real-time using gathered sensor data. As previously described, autonomous vehicles may rely on the accuracy of HD maps to safely drive. As such, if the HD map is inaccurate, the driving capabilities and safety of an autonomous vehicle may be compromised. Since the real world modeled by the HD map may change, the HD map would need to be updated as well to reflect the change in the real world.” …. [048] “The map-updating operation may begin at step 750, where the system may send sensor data and associated data gathered at that particular location to a server. In particular embodiments, the server may be associated with a transportation management system that manages a fleet of autonomous vehicles. In particular embodiments, the transmitted data may be the raw data obtained from the sensors or processed by-product data that represent the underlying raw data (the latter approach may have the advantage of reducing the amount of data transferred). The data transmitted to the server may be processed by the server to determine whether there is indeed a mismatch between the existing HD map and the current sensor measurement of the world. In particular embodiments, the server may also perform a comparison of the received data (and any objects detected therefrom) with a server-copy of the HD map to determine whether a mismatch exists. In particular embodiments, the server may also use data from other vehicles that have pass through the same location to determine whether the mismatch is consistent across data from different vehicles. In particular embodiments, the server may convert the data received from the vehicle into a latent representation, using the trained CNN described elsewhere herein, and compare the latent representation to a stored latent representation that reflects the HD map's current (not yet updated) model at that particular location. In particular embodiments, the server may transform the received data into generated map data, using the trained machine-learning model (e.g., CNN(s) and DCNN) described elsewhere herein, and compare the generated map data with the corresponding map data in the current HD map. If no mismatch is detected, the server may decide to not update the HD map. On the other hand, if the server determines that the current HD map does not include the detected object, the server may update the server-copy of the HD map as well as the local copies of the HD map on autonomous vehicles. In particular embodiments, the server may prioritize updating the HD maps of vehicles that would most likely be impacted by the HD map update. For example, the server may prioritize autonomous vehicles that are in the region (e.g., within a threshold distance) of where the new object is detected or have trajectories that would result in the vehicles being in that region in the near future.”
Drawings
The drawings are objected to because FIG 1 and FIG 2 both have elements that have numbers associated with them (Ex: 106, 102, 108, 112, 104, 110, 230, 224) but are not labeled as the other elements in the other figures. It would be appreciated if all elements were labeled for consistency. Examiner understands that the components in the two drawings are not the same components, and instead was indicating that textual descriptions accompany the numbered elements in figures 1 and 2 instead of having empty boxes with just numbers. Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. The figure or figure number of an amended drawing should not be labeled as “amended.” If a drawing figure is to be canceled, the appropriate figure must be removed from the replacement sheet, and where necessary, the remaining figures must be renumbered and appropriate changes made to the brief description of the several views of the drawings for consistency. Additional replacement sheets may be necessary to show the renumbering of the remaining figures. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 3-7, 9, 11-15, 17 and 19-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
101 Analysis – Step 1
Claim 1 is directed to a method, claim 9 is directed to a system and claim 17 is directed to one or more non-transitory computer-readable media. Therefore, claims 1, 9 and 17 are within at least one of the four statutory categories.
101 Analysis – Step 2A, Prong I
Regarding Prong I of the Step 2A analysis in the 2019 PEG, the claims are to be analyzed to determine whether they recite subject matter that falls within one of the follow groups of abstract ideas: a) mathematical concepts, b) certain methods of organizing human activity, and/or c) mental processes.
Independent claim 1 includes limitations that recite an abstract idea (emphasized below) and will be used as a representative claim for the remainder of the 101 rejection. The other analogous claims 9 and 17 are rejected for the same reasons as the representative claim 1 as discussed here. Claim 1 recites:
A method of generating high definition (HD) maps for a vehicle based on standard definition (SD) maps, the method comprising, using a processor:
receiving an SD map corresponding to an environment; receiving one or more aerial images corresponding to the environment;
predicting and labeling features contained within the SD map using at least one pairing of the SD map and a respective aerial image of the one or more aerial images;
generating, using a neural network, a first HD map based on the predicted and labeled features contained within the SD map and the one or more aerial images;
and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions;
and in real-time, while navigating the environment with the vehicle,
receiving images from one or more image sensors mounted on the vehicle,
and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
The examiner submits that the foregoing bolded limitation(s) constitute a “mental process” because under its broadest reasonable interpretation, the claim covers performance of the limitation in the human mind. For example, “predicting …”, “labeling …” and “generating …” all the various data in the context of this claim encompasses a person looking at data collected (received, detected, etc.) and forming a simple judgement (determination, analysis, comparison, etc.) either mentally or using a pen and paper. Accordingly, the claim recites at least one abstract idea. The Examiner notes that under MPEP 2106.04(a)(2)(III), the courts consider a mental process (thinking) that "can be performed in the human mind, or by a human using a pen and paper" to be an abstract idea. CyberSource Corp. v. Retail Decisions, Inc., 654 F.3d 1366, 1372, 99 USPQ2d 1690, 1695 (Fed. Cir. 2011). As the Federal Circuit explained, "methods which can be performed mentally, or which are the equivalent of human mental work, are unpatentable abstract ideas the ‘basic tools of scientific and technological work’ that are open to all.’" 654 F.3d at 1371, 99 USPQ2d at 1694 (citing Gottschalk v. Benson, 409 U.S. 63, 175 USPQ 673 (1972)). See also Mayo Collaborative Servs. v. Prometheus Labs. Inc., 566 U.S. 66, 71, 101 USPQ2d 1961, 1965 ("‘[M]ental processes[] and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work’" (quoting Benson, 409 U.S. at 67, 175 USPQ at 675)); Parker v. Flook, 437 U.S. 584, 589, 198 USPQ 193, 197 (1978) (same).
101 Analysis – Step 2A, Prong II
Regarding Prong II of the Step 2A analysis in the 2019 PEG, the claims are to be analyzed to determine whether the claim, as a whole, integrates the abstract into a practical application. As noted in the 2019 PEG, it must be determined whether any additional elements in the claim beyond the abstract idea integrate the exception into a practical application in a manner that imposes a meaningful limit on the judicial exception. The courts have indicated that additional elements merely using a computer to implement an abstract idea, adding insignificant extra solution activity, or generally linking use of a judicial exception to a particular technological environment or field of use do not integrate a judicial exception into a “practical application.”
In the present case, the additional limitations beyond the above-noted abstract idea are as follows (where the underlined portions are the “additional limitations” while the bolded portions continue to represent the “abstract idea”):
A method of generating high definition (HD) maps for a vehicle based on standard definition (SD) maps, the method comprising, using a processor:
receiving an SD map corresponding to an environment; receiving one or more aerial images corresponding to the environment;
predicting and labeling features contained within the SD map using at least one pairing of the SD map and a respective aerial image of the one or more aerial images;
generating, using a neural network, a first HD map based on the predicted and labeled features contained within the SD map and the one or more aerial images;
and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions;
and in real-time, while navigating the environment with the vehicle,
receiving images from one or more image sensors mounted on the vehicle,
and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
For the following reason(s), the examiner submits that the above identified additional limitations do not integrate the above-noted abstract idea into a practical application.
Regarding the additional limitations above, the examiner submits that these limitations are insignificant extra-solution activities that merely use a computer (processor) to perform the process. In particular, the receiving step is recited at a high level of generality (i.e. as a general means of receiving information for use in the predicting and other steps), and amounts to mere data gathering, which is a form of insignificant extra-solution activity. The inclusion that sensor data is obtained in “real-time” also does not impact this as sensors work in this manner and this received information is merely used to generate an additional HD map. The transmitting step is also recited at a high level of generality and amounts to mere post solution action, which is a form of insignificant extra-solution activity. Lastly, claims 9 and 17 further recite “A system configured to generate high definition (HD) maps for a vehicle based on standard definition (SD) maps, the system comprising: a processor” and “A non-tangible computer readable medium storing instructions that, when executed by a processor, cause the processor to generate high-definition (HD) maps for a vehicle based on standard definition (SD) maps, wherein executing the instructions causes the processor to” merely describes how to generally “apply” the otherwise mental judgements in a generic or general purpose mapping environment. See Alice Corp. Pty. Ltd. v. CLS Bank Int'l, 573 U.S. at 223 (“[T]he mere recitation of a generic computer cannot transform a patent-ineligible abstract idea into a patent-eligible invention.”). The device(s) and processor(s) are recited at a high level of generality and merely automates the steps.
Thus, taken alone, the additional elements do not integrate the abstract idea into a practical application. Further, looking at the additional limitation(s) as an ordered combination or as a whole, the limitation(s) add nothing that is not already present when looking at the elements taken individually. For instance, there is no indication that the additional elements, when considered as a whole, reflect an improvement in the functioning of a computer or an improvement to another technology or technical field, apply or use the above-noted judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition, implement/use the above-noted judicial exception with a particular machine or manufacture that is integral to the claim, effect a transformation or reduction of a particular article to a different state or thing, or apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is not more than a drafting effort designed to monopolize the exception (MPEP § 2106.05). Accordingly, the additional limitation(s) do/does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
101 Analysis – Step 2B
Regarding Step 2B of the 2019 PEG, as discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a processor to perform the steps amounts to nothing more than applying the exception using a generic computer component. Generally applying an exception using a generic computer component cannot provide an inventive concept. Furthermore, the addition of the “neural network” as an additional element is not sufficient to claim a practical application or amount to significantly more than a judicial exception as the limitation represents no more than mere instructions to apply the judicial exception on a computer and it can also be viewed as nothing more than an attempt to generally link the use of the judicial exception to the technological environment of computers. And as discussed above, the additional limitations discussed above are insignificant extra-solution activities.
The additional limitation of receiving information is a well-understood, routine and conventional activities because the background recites that the information can be from online/remote sources, and the specification does not provide any indication that the processor is anything other than a conventional computer. MPEP 2106.05(d)(II), and the cases cited therein, including Intellectual Ventures I, LLC v. Symantec Corp., 838 F.3d 1307, 1321 (Fed. Cir. 2016), TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610 (Fed. Cir. 2016), and OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363 (Fed. Cir. 2015), indicate that mere collection or receipt of data over a network is a well‐understood, routine, and conventional function when it is claimed in a merely generic manner. The additional limitation of “generating a first HD map …,” is a well-understood, routine, and conventional activity because the Federal Circuit in Trading Techs. Int’l v. IBG LLC, 921 F.3d 1084, 1093 (Fed. Cir. 2019), and Intellectual Ventures I LLC v. Erie Indemnity Co., 850 F.3d 1315, 1331 (Fed. Cir. 2017), for example, indicated that the mere performance which in the instant application is creating a map is a well understood, routine, and conventional function. Hence, the claim is not patent eligible.
Dependent claims 3-7, 11-15 and 19-20 do not recite any further limitations that cause the claims to be patent eligible. Rather, the limitations of dependent claims are directed toward additional aspects of the judicial exception and/or additional elements that do not integrate the judicial exception into a practical application. The dependent claims merely define limitations further or have additional steps such as “generating” and “receiving”. In addition, the use of a machine learning model merely confines the use of the abstract idea to a particular technological environment (neural networks) and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h). Therefore, dependent claims 2-7, 10-15 and 18-20 are not patent eligible.
Therefore, claims 1, 3-7, 9, 11-15, 17 and 19-20 are ineligible under 35 USC §101.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1, 3, 9, 11 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Buslaev (US 20250012600 A1) in view of Wang (US 20230349716 A1) in view of Arditi (US 20190147331 A1).
Regarding Claim 1, Buslaev teaches A method of generating high definition (HD) maps for a vehicle based on standard definition (SD) maps, the method comprising, using a processor (see at least [¶013, 018 and Claim 1]):
receiving one or more aerial images corresponding to the environment (Receiving aerial images corresponding to the environment. see at least [¶030, 034 & 041]);
predicting and labeling features contained within the SD map using at least one pairing of the SD map and a respective aerial image of the one or more aerial images (Predicting and labeling features from the SD map with a respective pairing of an aerial image from a plurality of aerial images available. Features that may be present in an aerial image and missing in an SD map are added to the SD map to have a more complete map. see at least [¶037-038, 040-041, 049-050 and 061-068]);
generating, using a neural network, a first HD map based on the predicted and labeled features contained within the SD map and the one or more aerial images (Generating, using a neural network, an HD map based on the predicted and labeled features contained within the SD map and aerial images. The neural network can be used to create the HD map elements used to generate the HD map from the additional data sources. see at least [¶037-042, 049-050 and 061-068]);
Buslaev does not explicitly teach receiving an SD map corresponding to an environment.
Shall be noted that Buslaev teaches having a map datastore that stores SD maps and would be assumed that the mapping system receives the SD map from this datastore. (see at least [“[0037] Moreover, the map datastore 330 can include both SD map data (e.g., map or map layers) and separate HD map data layers. Accordingly, the map datastore 330 may include a SD map datastore 332 and an HD map datastore 334.”] For more clarification the examiner is using secondary reference of Wang.
However, Wang does teach receiving an SD map corresponding to an environment (Receiving SD map corresponding to an environment that needs an HD map. see at least [¶032-035]).
Wang would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev to use the technique of receiving an SD map corresponding to an environment as taught by Wang. Doing so would lead to improved generation of an HD map from an SD map and additional data (see at least [¶026]).
Buslaev and Wang do not explicitly teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions; and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
However, Arditi does teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions (Transmitting an HD map to a vehicle for use in controlling see at least [¶014, 042, 052 & 060]).
and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle (In real-time while the vehicle is navigating an environment, receiving image sensor data from imaging sensors on the vehicle. see at least [¶043-045]),
and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map (Based on the received images, a CNN/DCNN1 and a previously existing HD map, generating an updated/second HD map. (CNN appear to already incorporate scene priors in their structure and thus would be used.) see at least [¶048-050 & 056-058]).
Arditi would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev and Wang to use the technique of transmitting the first HD map to the vehicle for use in controlling autonomous driving functions; and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map as taught by Arditi. Doing so would lead to improved accuracy in generated HD maps used for autonomous navigation (see at least [¶015]).
Regarding Claim 3 and 11, Buslaev, Wang and Arditi teach all of the limitations of claims 1 and 9 as shown above, Furthermore, Arditi teaches wherein generating the first HD map includes generating the first HD map using a large pre-trained neural network (LPNN) (Generating an HD map using trained neural networks (Neural networks are large in nature due to being trained with large datasets). see at least [¶014, 018-020 & 023]).
Arditi would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev and Wang to use the technique of having the first HD map is an offline map generated at a location remote from the vehicle as taught by Arditi. Doing so would lead to generation of HD maps for multiple areas using a single trained model (see at least [¶023]).
Regarding Claim 9, Buslaev teaches A system configured to generate high definition (HD) maps for a vehicle based on standard definition (SD) maps, the system comprising: a processor configured to (see at least [¶013, 018 and Claim 20]):
receiving one or more aerial images corresponding to the environment (Receiving aerial images corresponding to the environment. see at least [¶030, 034 & 041]);
predicting and labeling features contained within the SD map using at least one pairing of the SD map and a respective aerial image of the one or more aerial images (Predicting and labeling features from the SD map with a respective pairing of an aerial image from a plurality of aerial images available. Features that may be present in an aerial image and missing in an SD map are added to the SD map to have a more complete map. see at least [¶037-038, 040-041, 049-050 and 061-068]);
generating, using a neural network, a first HD map based on the predicted and labeled features contained within the SD map and the one or more aerial images (Generating, using a neural network, an HD map based on the predicted and labeled features contained within the SD map and aerial images. The neural network can be used to create the HD map elements used to generate the HD map from the additional data sources. see at least [¶037-042, 049-050 and 061-068]);
Buslaev does not explicitly teach receiving an SD map corresponding to an environment.
Shall be noted that Buslaev teaches having a map datastore that stores SD maps and would be assumed that the mapping system receives the SD map from this datastore. (see at least [“[0037] Moreover, the map datastore 330 can include both SD map data (e.g., map or map layers) and separate HD map data layers. Accordingly, the map datastore 330 may include a SD map datastore 332 and an HD map datastore 334.”] For more clarification the examiner is using secondary reference of Wang.
However, Wang does teach receiving an SD map corresponding to an environment (Receiving SD map corresponding to an environment that needs an HD map. see at least [¶032-035]).
Wang would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev to use the technique of receiving an SD map corresponding to an environment as taught by Wang. Doing so would lead to improved generation of an HD map from an SD map and additional data (see at least [¶026]).
Buslaev and Wang do not explicitly teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions.
Buslaev and Wang do not explicitly teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions; and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
However, Arditi does teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions (Transmitting an HD map to a vehicle for use in controlling see at least [¶014, 042, 052 & 060]).
and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle (In real-time while the vehicle is navigating an environment, receiving image sensor data from imaging sensors on the vehicle. see at least [¶043-045]),
and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map (Based on the received images, a CNN/DCNN2 and a previously existing HD map, generating an updated/second HD map. (CNN appear to already incorporate scene priors in their structure and thus would be used.) see at least [¶048-050 & 056-058]).
Arditi would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev and Wang to use the technique of transmitting the first HD map to the vehicle for use in controlling autonomous driving functions; and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map as taught by Arditi. Doing so would lead to improved accuracy in generated HD maps used for autonomous navigation (see at least [¶015]).
Regarding Claim 17, Buslaev teaches A non-tangible computer readable medium storing instructions that, when executed by a processor, cause the processor to generate high-definition (HD) maps for a vehicle based on standard definition (SD) maps, wherein executing the instructions causes the processor to (see at least [¶013, 018 and Claim 11]):
receiving one or more aerial images corresponding to the environment (Receiving aerial images corresponding to the environment. see at least [¶030, 034 & 041]);
predicting and labeling features contained within the SD map using at least one pairing of the SD map and a respective aerial image of the one or more aerial images (Predicting and labeling features from the SD map with a respective pairing of an aerial image from a plurality of aerial images available. Features that may be present in an aerial image and missing in an SD map are added to the SD map to have a more complete map. see at least [¶037-038, 040-041, 049-050 and 061-068]);
generating, using a neural network, a first HD map based on the predicted and labeled features contained within the SD map and the one or more aerial images (Generating, using a neural network, an HD map based on the predicted and labeled features contained within the SD map and aerial images. The neural network can be used to create the HD map elements used to generate the HD map from the additional data sources. see at least [¶037-042, 049-050 and 061-068]);
Buslaev does not explicitly teach receiving an SD map corresponding to an environment.
Shall be noted that Buslaev teaches having a map datastore that stores SD maps and would be assumed that the mapping system receives the SD map from this datastore. (see at least [“[0037] Moreover, the map datastore 330 can include both SD map data (e.g., map or map layers) and separate HD map data layers. Accordingly, the map datastore 330 may include a SD map datastore 332 and an HD map datastore 334.”] For more clarification the examiner is using secondary reference of Wang.
However, Wang does teach receiving an SD map corresponding to an environment (Receiving SD map corresponding to an environment that needs an HD map. see at least [¶032-035]).
Wang would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev to use the technique of receiving an SD map corresponding to an environment as taught by Wang. Doing so would lead to improved generation of an HD map from an SD map and additional data (see at least [¶026]).
Buslaev and Wang do not explicitly teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions.
Buslaev and Wang do not explicitly teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions; and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map.
However, Arditi does teach and transmitting the first HD map to the vehicle for use in controlling autonomous driving functions (Transmitting an HD map to a vehicle for use in controlling see at least [¶014, 042, 052 & 060]).
and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle (In real-time while the vehicle is navigating an environment, receiving image sensor data from imaging sensors on the vehicle. see at least [¶043-045]),
and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map (Based on the received images, a CNN/DCNN3 and a previously existing HD map, generating an updated/second HD map. (CNN appear to already incorporate scene priors in their structure and thus would be used.) see at least [¶048-050 & 056-058]).
Arditi would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev and Wang to use the technique of transmitting the first HD map to the vehicle for use in controlling autonomous driving functions; and in real-time, while navigating the environment with the vehicle, receiving images from one or more image sensors mounted on the vehicle, and based on the received images, one or more neural scene priors (NSPs), and the first HD map, generating a second HD map as taught by Arditi. Doing so would lead to improved accuracy in generated HD maps used for autonomous navigation (see at least [¶015]).
Claims 4-5, 12-13 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Buslaev (US 20250012600 A1) in view of Wang (US 20230349716 A1) in view of Arditi (US 20190147331 A1) in further view of Li (“Neural Scene Flow Prior”, 2021).
Regarding Claims 4 and 12, Buslaev, Wang and Arditi teach all of the limitations of claims 1 and 9 as shown above, Buslaev, Wang and Arditi do not explicitly teach wherein generating the first HD map includes generating the one or more NSPs and predicting and labeling the features based on the one or more NSPs.
However, Li does teach wherein generating the one or more NSPs and predicting and labeling the features based on the one or more NSPs (Generating offline maps using neural scene priors and being able to label/identify objects in the scene. see at least [Section 4.3, Page 7, Paragraph 4, Section 4.6, Page 8-9, All Paragraphs and Section 5, Page 9, Paragraph 3]).
Li would be in a similar field as it also deals in the area of map generation. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Wang and Arditi to use the technique of generating the one or more NSPs and predicting and labeling the features based on the one or more NSPs and predicting and labeling the features based on the one or more NSPs as taught by Li. Doing so would lead to improved estimating of 3D motion fields from large-scale dynamic scenes without annotating massive data but preserving generalizability (see at least [Broader Impact, Page 9, Paragraph 4]).
Regarding Claims 5 and 13, Buslaev, Wang, Arditi and Li teach all of the limitations of claims 4 and 12 as shown above, furthermore, Buslaev teaches wherein the features include at least one of lanes, line lines, centerlines or lanes, lane markings, traffic lights, traffic signs, and connections between the lanes (The features include lane lines/markings and traffic/road signs. see at least [¶037-038 & 041]).
Regarding Claim 19, Buslaev, Wang and Arditi teach all of the limitations of claim 17 as shown above, Furthermore, Arditi teaches wherein generating the first HD map includes generating the first HD map using a large pre-trained neural network (LPNN) (Generating an HD map using trained neural networks (Neural networks are large in nature due to being trained with large datasets). see at least [¶014, 018-020 & 023]).
Arditi would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev and Wang to use the technique of having the first HD map is an offline map generated at a location remote from the vehicle as taught by Arditi. Doing so would lead to generation of HD maps for multiple areas using a single trained model (see at least [¶023]).
Buslaev, Wang and Arditi do not explicitly teach wherein generate the one or more NSPs and predicting and labeling the features based on the one or more NSPs and predicting and labeling the features based on the one or more NSPs.
However, Li does teach wherein generate the one or more NSPs and predicting and labeling the features based on the one or more NSPs and predicting and labeling the features based on the one or more NSPs (Generating offline maps using neural scene priors and being able to label/identify objects in the scene. see at least [Section 4.3, Page 7, Paragraph 4, Section 4.6, Page 8-9, All Paragraphs and Section 5, Page 9, Paragraph 3]).
Li would be in a similar field as it also deals in the area of map generation. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Wang and Arditi to use the technique of generating the first HD map includes generating the one or more NSPs and predicting and labeling the features based on the one or more NSPs and predicting and labeling the features based on the one or more NSPs as taught by Li. Doing so would lead to improved estimating of 3D motion fields from large-scale dynamic scenes without annotating massive data but preserving generalizability (see at least [Broader Impact, Page 9, Paragraph 4]).
Claims 6, 14 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Buslaev (US 20250012600 A1) in view of Wang (US 20230349716 A1) in view of Arditi (US 20190147331 A1) in further view of Li (“Neural Scene Flow Prior”, 2021) in further view of Liu (“VectorMapNet: End-to-end Vectorized HD Map Learning”, 2023).
Regarding Claims 6 and 14, Buslaev, Wang, Arditi and Li teach all of the limitations of claim 5 and 13 as shown above, furthermore, Buslaev teaches further comprising generating the second HD map by, at the vehicle (Generating a second/updated HD map at the vehicle with new sensor information. see at least [¶059-060 & 072]):
generating perception data using the received images (Generating perception/raw data using the obtained images. see at least [¶067]);
Wang would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Arditi and Li to use the technique of generating the second HD map at the vehicle: generating perception data using the received images as taught by Wang. Doing so would lead to improved generation of an HD map from an SD map and additional data (see at least [¶026]).
Buslaev, Wang, Arditi and Li do not explicitly teach generating, using the perception data and the one or more NSPs, at least one of (i) a fused representation from sensor data and (ii) a bird’s eye view (BEV) of the environment; and generating the second HD map using the at least one of the fused representation and the BEV of the environment.
However, Liu does teach generating, using the perception data and the one or more NSPs, at least one of (i) a fused representation from sensor data and (ii) a bird’s eye view (BEV) of the environment (Generating using perception/raw sensor data and a CNN4 a fused representation from the sensor data and a BEV view of the environment. (CNN appear to already incorporate scene priors in their structure.) see at least [Section 3.2, Page 4, All Paragraphs and FIG 2]);
and generating the second HD map using the at least one of the fused representations and the BEV of the environment (Generating an HD map using the fused sensor representation and the BEV of the environment. see at least [Section 3.3, Page 5, Paragraph 1, Section 3.4, Page 5, Paragraph 1 and FIG 2]).
Liu would be in a similar field as it also deals in the area of HD map generation. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Wang, Arditi and Li to use the technique of generating, using the perception data and the one or more NSPs, at least one of (i) a fused representation from sensor data and (ii) a bird’s eye view (BEV) of the environment; and generating the second HD map using the at least one of the fused representation and the BEV of the environment as taught by Liu. Doing so would lead to an improved HD map generation using polylines primitives derived from sensor data (see at least [Conclusion, Page 9]).
Regarding Claim 20, Buslaev, Wang, Arditi and Li teach all of the limitations of claim 19 as shown above, furthermore, Buslaev teaches further comprising generating the second HD map by, at the vehicle (Generating a second/updated HD map at the vehicle with new sensor information. see at least [¶059-060 & 072]):
generating perception data using the received images (Generating perception/raw data using the obtained images. see at least [¶067]);
Wang would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Arditi and Li to use the technique of generating the second HD map at the vehicle: generating perception data using the received images as taught by Wang. Doing so would lead to improved generation of an HD map from an SD map and additional data (see at least [¶026]).
Buslaev, Wang, Arditi and Li do not explicitly teach generating, using the perception data and the one or more NSPs, at least one of (i) a fused representation from sensor data and (ii) a bird’s eye view (BEV) of the environment; and generating the second HD map using the at least one of the fused representation and the BEV of the environment.
However, Liu does teach generating, using the perception data and the one or more NSPs, at least one of (i) a fused representation from sensor data and (ii) a bird’s eye view (BEV) of the environment (Generating using perception/raw sensor data and a CNN5 a fused representation from the sensor data and a BEV view of the environment. (CNN appear to already incorporate scene priors in their structure.) see at least [Section 3.2, Page 4, All Paragraphs and FIG 2]);
and generating the second HD map using the at least one of the fused representations and the BEV of the environment (Generating an HD map using the fused sensor representation and the BEV of the environment. see at least [Section 3.3, Page 5, Paragraph 1, Section 3.4, Page 5, Paragraph 1 and FIG 2]).
Liu would be in a similar field as it also deals in the area of HD map generation. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Wang, Arditi and Li to use the technique of generating, using the perception data and the one or more NSPs, at least one of (i) a fused representation from sensor data and (ii) a bird’s eye view (BEV) of the environment; and generating the second HD map using the at least one of the fused representation and the BEV of the environment as taught by Liu. Doing so would lead to an improved HD map generation using polylines primitives derived from sensor data (see at least [Conclusion, Page 9]).
Claims 7-8 and 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Buslaev (US 20250012600 A1) in view of Wang (US 20230349716 A1) in view of Arditi (US 20190147331 A1) in further view of Li (“Neural Scene Flow Prior”, 2021) in further view of Liu (“VectorMapNet: End-to-end Vectorized HD Map Learning”, 2023) in further view of Wray (US 20220306156 A1).
Regarding Claims 7 and 15, Buslaev, Wang, Arditi, Li and Liu teach all of the limitations of claims 6 and 14 as shown above, Buslaev, Wang, Arditi, Li and Liu do not explicitly teach further comprising generating a lane-level trajectory associated with a planned route for the vehicle using the second HD map.
However, Wray does teach further comprising generating a lane-level trajectory associated with a planned route for the vehicle using the second HD map (Generating a lane-level trajectory for a planned route using a generated HD map. HD maps can be updated/generated with additional data. see at least [¶046-047, 0214 & 0224]).
Wray would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Wang, Arditi, Li and Liu to use the technique generating a lane-level trajectory associated with a planned route for the vehicle using the second HD map as taught by Wray. Doing so would lead to allowing vehicles to navigate autonomously using generated HD maps (see at least [¶046]).
Regarding Claims 8 and 16, Buslaev, Wang, Arditi, Li and Liu teach all of the limitations of claims 7 and 15 as shown above, furthermore, Wray Teaches further comprising executing autonomous driving commands to autonomously navigate the vehicle based on the lane-level trajectory and the second HD map (Executing autonomous driving based on the lane-level trajectory and the HD map. HD maps can be updated/generated with additional data. see at least [¶045-047, 0214 & 0224]).
Wray would be in a similar field as it also deals in the area of generating HD maps. Therefore, it would have been obvious to those having ordinary skill in the art before the effective filing date of the instant application to modify Buslaev, Wang, Arditi, Li and Liu to use the technique executing autonomous driving commands to autonomously navigate the vehicle based on the lane-level trajectory and the second HD map as taught by Wray. Doing so would lead to allowing vehicles to navigate autonomously using generated HD maps (see at least [¶046]).
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner
should be directed to MOISES GASCA ALVA JR whose telephone number is (571)272-3752. The examiner
can normally be reached Monday-Friday 6:30 - 4:00. 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,
Faris Almatrahi can be reached on (313) 446-4821. 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.
/MOISES GASCA ALVA/Examiner, Art Unit 3667
/FARIS S ALMATRAHI/Supervisory Patent Examiner, Art Unit 3667
1 The superior performance of CNNs on computer vision tasks can be attributed to their structure, tailored to image data. A recent paper, called Deep Image Prior, shows that a CNN’s architecture already incorporates the prior for natural scene statistics. Unlike fully connected networks, CNNs are able to take advantage of the fact that relationships among close pixels matter more than among those far apart. - https://medium.com/data-science/cnn-cheat-sheet-the-essential-summary-for-a-quick-start-58820a14d3b4
2 The superior performance of CNNs on computer vision tasks can be attributed to their structure, tailored to image data. A recent paper, called Deep Image Prior, shows that a CNN’s architecture already incorporates the prior for natural scene statistics. Unlike fully connected networks, CNNs are able to take advantage of the fact that relationships among close pixels matter more than among those far apart. - https://medium.com/data-science/cnn-cheat-sheet-the-essential-summary-for-a-quick-start-58820a14d3b4
3 The superior performance of CNNs on computer vision tasks can be attributed to their structure, tailored to image data. A recent paper, called Deep Image Prior, shows that a CNN’s architecture already incorporates the prior for natural scene statistics. Unlike fully connected networks, CNNs are able to take advantage of the fact that relationships among close pixels matter more than among those far apart. - https://medium.com/data-science/cnn-cheat-sheet-the-essential-summary-for-a-quick-start-58820a14d3b4
4 The superior performance of CNNs on computer vision tasks can be attributed to their structure, tailored to image data. A recent paper, called Deep Image Prior, shows that a CNN’s architecture already incorporates the prior for natural scene statistics. Unlike fully connected networks, CNNs are able to take advantage of the fact that relationships among close pixels matter more than among those far apart. - https://medium.com/data-science/cnn-cheat-sheet-the-essential-summary-for-a-quick-start-58820a14d3b4
5 The superior performance of CNNs on computer vision tasks can be attributed to their structure, tailored to image data. A recent paper, called Deep Image Prior, shows that a CNN’s architecture already incorporates the prior for natural scene statistics. Unlike fully connected networks, CNNs are able to take advantage of the fact that relationships among close pixels matter more than among those far apart. - https://medium.com/data-science/cnn-cheat-sheet-the-essential-summary-for-a-quick-start-58820a14d3b4
Prosecution Insights