Prosecution Insights
Last updated: November 20, 2026
Application No. 18/799,041

DEVICE AND COMPUTER-IMPLEMENTED METHOD FOR MACHINE LEARNING

Final Rejection §101§102§103
Filed
Aug 09, 2024
Priority
Sep 06, 2023 — DE 10 2023 208 618.2
Examiner
ZHAO, LEI
Art Unit
2668
Tech Center
2600 — Communications
Assignee
Robert Bosch GmbH
OA Round
2 (Final)
74%
Grant Probability
Favorable
3-4
OA Rounds
9m
Est. Remaining
93%
With Interview

Examiner Intelligence

Grants 74% — above average
74%
Career Allowance Rate
57 granted / 77 resolved
+12.0% vs TC avg
Strong +19% interview lift
Without
With
+18.8%
Interview Lift
resolved cases with interview
Typical timeline
3y 0m
Avg Prosecution
22 currently pending
Career history
110
Total Applications
across all art units

Statute-Specific Performance

§101
6.8%
-33.2% vs TC avg
§103
64.0%
+24.0% vs TC avg
§102
26.5%
-13.5% vs TC avg
§112
2.7%
-37.3% vs TC avg
Black line = Tech Center average estimate • Based on career data from 77 resolved cases

Office Action

§101 §102 §103
CTNF 18/799,041 CTNF 98850 Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Claim Rejections - 35 USC § 101 07-04-01 AIA 07-04 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. Claim 9 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because claim 9 recites “A computer-readable medium”. A computer-readable medium as recited is not patent eligible subject matter because it can be “a carrier wave”. “Thus, a claim to a computer readable medium that can be a compact disc or a carrier wave covers a non-statutory embodiment and therefore should be rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.” MPEP 2106.03(II) . A recommended remedy for claiming a computer readable medium is to have it embodied within a “non-transitory” computer readable medium. See also USPTO Published 2019 Patent Eligibility Guidance, and MPEP 2106 and 2106.03 for additional guidance . Claim Rejections - 35 USC § 102 07-07-aia AIA 07-07 The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action: A person shall be entitled to a patent unless – 07-08-aia AIA (a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention. 07-12-aia AIA (a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention. 07-15 AIA Claim s 1-3, 6 and 8-9 are rejected under 35 U.S.C. 102( a)(1 ) as being anticipated by Yang (LOGICDEF: An Interpretable Defense Framework against Adversarial Examples via Inductive Scene Graph Reasoning, The Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI-22)) . Regarding claim 1, Yang teaches a computer-implemented method for machine learning (We show that LOGICDEF can incorporate knowledge and taxonomy from ConceptNet to obtain additional improvements with curriculum learning . Page 2 left column 5 th paragraph) , comprising the following steps: providing a first class and a second class of a classification (For example, consider an image of a train as shown in Figure 1. The “ train ” (i.e., a first class of a classification) object in the image is classified as a “ cat ” (i.e., a second class of a classification) by the deep object classifier due to the adversarial attack. Page 8840 right column last paragraph) ; providing a first set of scene graphs, wherein the first set of scene graphs includes scene graphs of digital images that are incorrectly classified in the first class or the second class (For example, consider an image of a train as shown in Figure 1. The “train” object in the image is classified as a “cat” by the deep object classifier due to the adversarial attack. Page 8840 right column last paragraph. PNG media_image1.png 608 700 media_image1.png Greyscale ) ; providing a second set of scene graphs, wherein the second set of scene graphs includes scene graphs of digital images that are correctly classified with respect to the first class or the second class (Furthermore, the bounded object has components “door” and “window”, and is above “tracks”, and these relationships are more likely associated with “train”. Page 8840 right column last paragraph. PNG media_image2.png 608 700 media_image2.png Greyscale ) ; and determining, depending on the first set of scene graphs (which reads on “Scene Graph generated in Training Phase” in Figure 2) and the second set of scene graphs (which reads on “Scene Graph generated in Attack Phase” in Figure 2) , a rule ( PNG media_image3.png 400 970 media_image3.png Greyscale ) that indicates that: (i) a presence of a first object and/or a second object in a digital image (We instead aim to defend against adversarial patch attacks on an object classifier, that is, the attack targets the classification of a single object in the scene (e.g. the “train” in Figure 1). Page 8841 right column 2 nd paragraph) , and/or (ii) a relation between the first object and the second object in a scene graph of the digital image (The claim language is interpreted as disjunctive.) , results in that a classification of the digital image includes a misclassification of the digital image into the second class instead of the first class (For example, consider an image of a train as shown in Figure 1. The “ train ” object in the image is classified as a “ cat ” by the deep object classifier due to the adversarial attack. Page 8840 right column last paragraph) . Regarding claim 2, Yang teaches the method according to claim 1, wherein the digital image is captured by a sensor (However, recent studies have demonstrated that these models are vulnerable to adversarial attack, exposing a risk-of-use in critical applications where untrusted parties have access to the data environment or even directly to the sensor inputs. Abstract) , including a camera (For example, consider an image of a train as shown in Figure 1. Page 8840 right column last paragraph. PNG media_image4.png 608 700 media_image4.png Greyscale ) , or a radar sensor, or a lidar sensor, or an infrared sensor, or an ultrasound sensor, or a motion sensor. Regarding claim 3, Yang teaches the method according to claim 1, further comprising: (i) detecting an anomaly depending on the rule ( PNG media_image5.png 88 614 media_image5.png Greyscale . Page 8843 right column 1 st paragraph) , or (ii) determining a cause of an anomaly depending on the rule. Regarding claim 6, Yang teaches the method according to claim 1, wherein the method further comprises determining a logic program depending on the first set of scene graphs and the second set of scene graphs (We show defense examples by LOGICDEF-AUTO in Figure 3. The left column shows the patched images and the objects that are attacked. The middle column shows the scene graphs and the logic rules that are used for defenses. For the right column, the first row shows the probability output of the classifier on a clean (which are the results obtained based on the correct scene graphs, i.e., the second set of scene graphs) and patched image (which are the results obtained based on the incorrect scene graphs, i.e., the first set of scene graphs) respectively and the second row shows the probability output of the logic rules and the final defense model, respectively. Page 8846 left column 2 nd paragraph. PNG media_image6.png 698 1292 media_image6.png Greyscale ) , and determining the rule depending on the logic program ( PNG media_image7.png 698 1292 media_image7.png Greyscale ) . Apparatus claim 8 is drawn to the apparatus corresponding to the method of using same as claimed in claim 1. Therefore apparatus claim 8 corresponds to method claim 1, and is rejected for the same reasons of anticipation as used above. Claim 9 is drawn to a computer-readable medium having executable instructions stored for carrying out the method of claim 1. Therefore, claim 9 corresponds to method claim 1, and is rejected for the same reasons of anticipation as used above . Claim Rejections - 35 USC § 103 07-20-aia AIA 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. 07-21-aia AIA Claim s 4-5 are rejected under 35 U.S.C. 103 as being unpatentable over Yang (LOGICDEF: An Interpretable Defense Framework against Adversarial Examples via Inductive Scene Graph Reasoning, The Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI-22)) hereinafter Yang, in view of Peng (Chinese Patent Pub. No.: CN 114863394 A), hereinafter Peng . Regarding claim 4, Yang teaches all of the elements of the claimed invention as stated in claim 3 except for the following limitations as further recited. However, Peng teaches further comprising: operating a technical system depending on the classification (For example, adversarial attacks on medical IoT systems can lead to misclassification of computed tomography (CT) images, increasing the risk of misdiagnosis. [n0098]) ; and operating the technical system independent of the classification in case an anomaly is detected depending on the rule (Therefore, IoT systems need to be highly robust and able to effectively identify abnormal data so that they can make reliable decisions based on correct data and improve the security of the IoT system. [n0099]) ; wherein the technical system includes a surveillance system or a medical imaging system (For example, adversarial attacks on medical IoT systems can lead to misclassification of computed tomography (CT) images, increasing the risk of misdiagnosis. [n0098]) . It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Yang to incorporate the teachings of Peng to operate a surveillance or a medical imaging system depending on the classification when no anomaly is detected and independent of the classification in case an anomaly is detected depending on the rule so that they can make reliable decisions based on correct data and improve the security of the system. Regarding claim 5, Peng in the combination teaches the method according to claim 4, wherein the operating of the technical system includes: when no anomaly is detected, outputting an output including a control signal for controlling the technical system, or a display of the classification for displaying by the technical system, the output being determined depending on the classification (For example, adversarial attacks on medical IoT systems can lead to misclassification of computed tomography (CT) images, increasing the risk of misdiagnosis. [n0098]. In other words, the medical IoT system outputs the classification when no anomaly is detected.) ; and when the anomaly is detected, not outputting the output, or outputting the output with an indication that the anomaly is detected (It is evident that electronic devices can perform consistency checks based on the scene semantics described by the sensor network (i.e., compare similarity with the semantic centroids of each scene), effectively detect whether there are abnormal sensors or whether the sensor system has been intruded in the sensor network, trace the source of abnormal sensors, and issue alarms . [n0204]) . 07-21-aia AIA Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Yang (LOGICDEF: An Interpretable Defense Framework against Adversarial Examples via Inductive Scene Graph Reasoning, The Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI-22)) hereinafter Yang, in view of Chen (Consistent Scene Graph Generation by Constraint Optimization, ASE ’22, October 10–14, 2022, Rochester, MI, USA), hereinafter Chen . Regarding claim 7, Yang teaches all of the elements of the claimed invention as stated in claim 1 except for expressly teaching the following limitations as further recited. However, Chen teaches wherein the method further comprises: providing a label for at least one digital image (For the object type, we set a constraint that each scene must have a person-like object (i.e. " Person", "Man" or "Woman "). Page 10 left column 6 th paragraph) ; determining a classification for the at least one digital image (To simplify the comparison, we perform scene graph classification where the ground truth object bounding box is provided, and the task is to find the object and relation types. Page 10 left column last paragraph) , depending on at least a part of the at least one digital image including pixel values of the at least one digital image (During training, the object detector is trained first with the training images while using object masks (i.e. a binary matrix where 1 means that a particular pixel is part of the object) as ground truth. Page 2 right column 7 th paragraph) ; determining a scene graph for the at least one digital image (During operation (i.e. actual scene graph generation ), the prediction is similar to training: (1) The object detector predicts the object masks. (2) The masks are fed to the attribute and relation extractor to predict the output scene graph. (3) Since the predicted attributes and relations are probability distributions over the possible values, a probabilistic scene graph is first obtained as the output. (4) The final scene graph 𝐺 𝑚𝑙 is chosen by maximum likelihood decision, i.e., by selecting the most probable value individually for each attribute and relation prediction. Page 2 right column last paragraph) , depending on at least a part of the at least one digital image including pixel values of the at least one digital image (During training, the object detector is trained first with the training images while using object masks (i.e. a binary matrix where 1 means that a particular pixel is part of the object) as ground truth. Page 2 right column 7 th paragraph) ; determining, depending on the label and the classification of the at least one digital image, whether the classification of the at least one digital image is correct or not (Scene Accuracy (SA) measures the proportion of predicted scenes that completely match a ground truth scene. Page 4 left column 4 th paragraph) . It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Yang to incorporate the teachings of Chen to determine a scene graph, depending on a part of the digital image including pixel values and determine, depending on the label and the classification, whether the classification of the digital image is correct or not in order to enhance the robustness of defense models. Yang in the combination further teaches adding the scene graph for the at least one digital image to the first set of scene graphs when the classification is incorrect, or adding the scene graph for the at least one digital image to the second set of scene graphs when the classification is correct (We show defense examples by LOGICDEF-AUTO in Figure 3. The left column shows the patched images and the objects that are attacked. The middle column shows the scene graphs and the logic rules that are used for defenses. For the right column, the first row shows the probability output of the classifier on a clean (i.e., when the classification is correct) and patched image respectively and the second row shows the probability output of the logic rules and the final defense model, respectively. Page 8846 left column 2 nd paragraph. PNG media_image8.png 698 1292 media_image8.png Greyscale ) . Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to LEI ZHAO whose telephone number is (703)756-1922. The examiner can normally be reached Monday - Friday 8:00 am - 5:00 pm. 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, VU LE can be reached at (571)272-7332. 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. /LEI ZHAO/Examiner, Art Unit 2668 /VU LE/Supervisory Patent Examiner, Art Unit 2668 Application/Control Number: 18/799,041 Page 2 Art Unit: 2668 Application/Control Number: 18/799,041 Page 3 Art Unit: 2668 Application/Control Number: 18/799,041 Page 4 Art Unit: 2668 Application/Control Number: 18/799,041 Page 5 Art Unit: 2668 Application/Control Number: 18/799,041 Page 6 Art Unit: 2668 Application/Control Number: 18/799,041 Page 7 Art Unit: 2668 Application/Control Number: 18/799,041 Page 8 Art Unit: 2668 Application/Control Number: 18/799,041 Page 9 Art Unit: 2668 Application/Control Number: 18/799,041 Page 10 Art Unit: 2668 Application/Control Number: 18/799,041 Page 11 Art Unit: 2668 Application/Control Number: 18/799,041 Page 12 Art Unit: 2668 Application/Control Number: 18/799,041 Page 13 Art Unit: 2668 Application/Control Number: 18/799,041 Page 14 Art Unit: 2668
Read full office action

Prosecution Timeline

Aug 09, 2024
Application Filed
Jun 01, 2026
Non-Final Rejection mailed — §101, §102, §103
Aug 31, 2026
Response Filed
Nov 18, 2026
Final Rejection mailed — §101, §102, §103 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12797159
APPARATUS AND METHOD FOR GENERATING VIRTUAL MODEL
4y 4m to grant Granted Nov 17, 2026
Patent 12802201
METHOD FOR INSPECTING A CORRECT EXECUTION OF A PROCESSING STEP OF COMPONENTS, IN PARTICULAR A WIRING HARNESS, DATA STRUCTURE, AND SYSTEM
3y 5m to grant Granted Nov 17, 2026
Patent 12793235
MOTION-BASED TARGET TRACKING
5y 2m to grant Granted Nov 10, 2026
Patent 12784245
Image Processing and Segmentation of Sets of Z-Stacked Images of Three-Dimensional Biological Samples
3y 4m to grant Granted Nov 03, 2026
Patent 12784280
ENCODING AND DECODING METHOD, APPARATUS, AND DEVICE, STORAGE MEDIUM, COMPUTER PROGRAM, AND COMPUTER PROGRAM PRODUCT
2y 11m to grant Granted Nov 03, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

3-4
Expected OA Rounds
74%
Grant Probability
93%
With Interview (+18.8%)
3y 0m (~9m remaining)
Median Time to Grant
Moderate
PTA Risk
Based on 77 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month