Prosecution Insights
Last updated: April 19, 2026
Application No. 18/619,612

INSPECTION OF MICROELECTRONICS USING NEURAL NETWORK

Non-Final OA §103
Filed
Mar 28, 2024
Examiner
HUYNH, THANG GIA
Art Unit
2611
Tech Center
2600 — Communications
Assignee
Rockwell Collins Inc.
OA Round
1 (Non-Final)
76%
Grant Probability
Favorable
1-2
OA Rounds
2y 4m
To Grant
99%
With Interview

Examiner Intelligence

Grants 76% — above average
76%
Career Allow Rate
19 granted / 25 resolved
+14.0% vs TC avg
Strong +50% interview lift
Without
With
+50.0%
Interview Lift
resolved cases with interview
Typical timeline
2y 4m
Avg Prosecution
21 currently pending
Career history
46
Total Applications
across all art units

Statute-Specific Performance

§101
2.3%
-37.7% vs TC avg
§103
73.9%
+33.9% vs TC avg
§102
7.7%
-32.3% vs TC avg
§112
11.5%
-28.5% vs TC avg
Black line = Tech Center average estimate • Based on career data from 25 resolved cases

Office Action

§103
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 . Claim Objections Claims 11-12 are objected to because of the following informalities: Regarding Claim 11, Claim 11 recites, “the at least one first color image patch and the at least one second color image patch” while its base Claim of 1 recites, “at least one first smaller color image patch . . . at least one second smaller color image patch”. Claim 11 should be corrected to have consistent terminology with its base Claim. Regarding Claim 12, Claim 12 recites, “the at least one first color patch or the at least one second color patch” while its base Claim of 1 recites, “at least one first smaller color image patch . . . at least one second smaller color image patch”. Claim 12 should be corrected to have consistent terminology with its base Claim. Appropriate correction is required. 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. This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention. Claims 1-6 and 13-17 are rejected under 35 U.S.C. 103 as being unpatentable over Kelly et al. (US11423577B2) (Hereinafter referred to as Kelly) in view of Zhang et al. (“Deep Learning Based Defect Detection for Solder Joints on Industrial X-Ray Circuit Board Images”) (Hereinafter referred to as Zhang) and in further view of Wang (US 11222234 B2). Regarding Claim 1, Kelly discloses A method for training a neural network model, the method comprising: (Col 6 Lines 21-23, “In some embodiments, the neural network 400 may be trained using stochastic gradient descent on generated training images.”) generate a first image stack, a second image stack, and a third image stack, wherein the first image stack, the second image stack, and the third image stack are two-dimensional (2D) x-ray images; (Col 1 Lines 14-23, “One conventional way of inspecting PCBAs involves the use of automated x-ray inspection equipment (AXI). The AXI typically captures 2-dimensional (2-d) gray scale images of interconnects which can then be analyzed using a simple linear regression model. If a defect is identified using the linear regression model and the 2-d images, then a second, offline inspection is performed using a 3-dimensional (3-d) x-ray image to verify the presence of a defect. During the second, offline inspection the PCBA is rotated to obtain the 3-d x-ray image.” Also see Fig. 2 showing example of three planes being used for generating image stacks.) converting the first image stack, the second image stack, and the third image stack into a respective greyscale image of each stack; (Col 1 Lines 16-17, “The AXI typically captures 2-dimensional (2-d) gray scale images . . .” Note that the end result is still grey scale images, and that the capturing process of the image can be considered as “converting” the image stack into a greyscale image as the captured images can start at a state which is not gray scale. Also note that the application specification simply expresses that “Each image stack is a two-dimensional (2D) x-ray image.” and typically, X-ray images are already in gray scale.) combining the first greyscale image stack, the second greyscale image stack, and the third greyscale image stack into a color image, wherein the color image is a 2D image; (Abstract, “The method further comprises converting the plurality of 2-dimensional gray scale images into a color image. Each of the plurality of 2-dimensional gray scale images corresponds to and is used as input for a respective color channel of the color image.” Col 3 Lines 51-60, “In this example, the x-ray device 102 provides the images 320-1, 320-2, and 320-N . . .The color image generator 110 generates a color image using the images 320-1, 320-2 and 320-N. In particular, the color image generator 110 is configured to use one of each of the images 320-1, 320-2, and 320-N as an input into a respective color channel of a color image 324, shown in FIG. 3.”) identifying within the color image as including at least one anomalous solder ball; and (Col 4 Lines 9-18, “The color image analyzer 108 is configured to identify defects, such as example defect 326, based on analysis of the color image 324. In particular, the color image analyzer 108 implements machine learning techniques to identify defects in the color image 324. The machine learning techniques can include algorithms or models that are generated by performing supervised, unsupervised, or semi-supervised learning on a dataset, and subsequently applying the generated algorithm or model to identify defects, such as the head-in-pillow defect discussed above.” Col 10 Lines 22-24, “For example, in some embodiments, the color variation can indicate a defect in a solder ball connection of a ball grid array.”) training the neural network model, wherein the neural network model is trained to predict an anomaly within the 3D x-ray model of the plurality of BGA solder balls. (Col 4 Lines 9-18, “The color image analyzer 108 is configured to identify defects, such as example defect 326, based on analysis of the color image 324. In particular, the color image analyzer 108 implements machine learning techniques to identify defects in the color image 324. The machine learning techniques can include algorithms or models that are generated by performing supervised, unsupervised, or semi-supervised learning on a dataset, and subsequently applying the generated algorithm or model to identify defects, such as the head-in-pillow defect discussed above”) However, Kelly fails to explicitly disclose discretizing a three-dimensional (3D) x-ray model of a plurality of ball grid array (BGA) solder balls to generate a first image stack, a second image stack, and a third image stack, wherein the first image stack, the second image stack, and the third image stack are two-dimensional (2D) x-ray images; . . . determining at least one solder ball as anomalous within at least one of the first greyscale image stack, the second greyscale image stack, and the third greyscale image stack; . . . identifying at least one first smaller color image patch within the color image as including at least one anomalous solder ball and at least one second smaller color image patch within the color image as including at least one normal solder ball; and training the neural network model with a set of color image patches including the at least one first smaller color image patch having the at least one anomalous solder ball and the at least one second smaller color image patch having the at least one normal solder ball, wherein the neural network model is trained to predict an anomaly within the 3D x-ray model of the plurality of BGA solder balls. Zhang teaches discretizing a three-dimensional (3D) x-ray model of a plurality of ball grid array (BGA) solder balls to generate a first image stack, a second image stack, and a third image stack, wherein the first image stack, the second image stack, and the third image stack are two-dimensional (2D) x-ray images; (Page 3 Fig. 2 “An example of 3D X-ray imaging with four slices.” In combination with Kelly, instead of four slices, the 3D X-ray would be discretized into the three image stacks based on the slice planes shown in Kelly Fig. 2.) identifying at least one first smaller color image patch within the color image as including at least one anomalous solder ball and at least one second smaller color image patch within the color image as including at least one normal solder ball; and (Page 2 Left Column Paragraph 2, “Recently, deep learning such as convolutional neural networks (CNNs) have shown outstanding performance on image based tasks such as image classification and object detection. . .They incorporate the whole image as well as region of interest patches to classify whether a joint is qualified or unqualified. . . It requires manual labeling not only for the whole image, but also for each image, four to five patches are required manual labeling.” In this case, classifying whether a joint is qualified or unqualified corresponds to identifying the at least one anomalous solder ball and the at least one normal solder ball, and in combination with Kelly, the region of interest patches would be based on the color image and thus “at least one first smaller color image patch within the color image as including at least one anomalous solder ball and at least one second smaller color image patch within the color image as including at least one normal solder ball”.) training the neural network model with a set of color image patches including the at least one first smaller color image patch having the at least one anomalous solder ball and the at least one second smaller color image patch having the at least one normal solder ball, wherein the neural network model is trained to predict an anomaly within the 3D x-ray model of the plurality of BGA solder balls. (In combination with Kelly already teaching training the neural network with a dataset in a supervised manner, and Zhang Page 2 Left Column Paragraph 2 teaching the image patches and classify abnormal and normal, the above limitation of training a neural network with those specific color patches are taught.) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Kelly with Zhang to include a 3D-Xray model and discretizing it, as well as image patches and detecting qualified and unqualified patches. The motivation to combine would have been Kelly with Zhang would have been obvious as both arts are within the same field of identifying solder defects (See Zhang Abstract). The benefit of using image patches with anomalous and normal solder ball for training the neural network is that it allows the network to learn and classify these balls within the image. However, Kelly in view of Zhang still fail to explicitly disclose determining at least one solder ball as anomalous within at least one of the first greyscale image stack, the second greyscale image stack, and the third greyscale image stack; Wang teaches determining at least one solder ball as anomalous within at least one of the first greyscale image stack, the second greyscale image stack, and the third greyscale image stack; (Col 1 Lines 45-51, “a method of training a convolutional neural network through deep learning for defect inspection. The method includes collecting a training sample set including multiple solder joint images. A respective one of the multiple solder joint images includes at least one of multiple solder joints having different types of solder joint defects.” Col 5 Lines 10-14, “Optionally, the imaging system includes a radiography imager including a radiation source, a sample desk, and a detector device. The radiation source includes one selected from X-ray source, γ-ray source, e-beam source.” Col 6 Lines 55-60, “For example, defect categories such as . . . Some flaws (solder ball, cutting-off, etc.) may be related to any of copper wires, solder pads, and solder joints.” In this case, Kelly already teaches grayscale image stacks, and in combination with Wang teaching defect inspection of X-ray images using a neural network, the above limitation of detecting at least one solder ball anomalous within the greyscale images is taught.) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Kelly in view of Zhang with Wang to include detecting at least one solder ball as anomalous within the greyscale image stacks. The motivation to combine Kelly in view of Zhang with Wang would have been obvious as all three arts are within the same field of detecting solder defects (See Wang Abstract). One possible benefit of detecting anomalous solder ball within the greyscale image stacks is that it would allow the neural network work on detecting defects that are not color based. Regarding Claim 2, Kelly in view of Zhang and Wang disclose The method of claim 1, wherein the first image stack, the second image stack, and the third image stack are generated from a common reference point. (See Kelly Fig. 2 showing example of the three planes used for generating image stacks. The planes can be considered as a common reference point.) Regarding Claim 3, Kelly in view of Zhang and Wang disclose The method of claim 1, wherein discretizing the 3D x-ray model includes generating the first image stack from a plane within the 3D x-ray model of the BGA solder joints just above a pad connected to the BGA solder joints. (See Kelly Fig. 2 showing pad 209. As shown by the image, one can have the location of a first plane be just above the pad 209.) Regarding Claim 4, Kelly in view of Zhang and Wang disclose The method of claim 3, further comprising generating the second image stack from a plane halfway between a top of the BGA solder joints and the plane of the first image stack. (See Kelly Fig. 2 showing parallel planes on a solder ball. Also see Kelly Col 3 Lines 36-41, “The example of FIG. 2 includes a portion of ball grid array 201, solder ball 205, solder paste 207, pad 209, and printed circuit board 211. The x-ray device 102 is configured to capture a respective 2-d gray scale image at each of planes 203-1, 203-2, and 203-N, where N is the total number of planes.” Note that since Kelly doesn’t specify where the planes are have to be located, one could easily have the second image stack be from a plane halfway between a top of the BGA solder joints and the plane of the first image stack.) Regarding Claim 5, Kelly in view of Zhang and Wang disclose The method of claim 4, further comprising generating the third image stack from a plane halfway between the plane for the first image stack and the plane for the second image stack. (See Kelly Fig. 2 showing parallel planes on a solder ball. Also see Kelly Col 3 Lines 36-41, “The example of FIG. 2 includes a portion of ball grid array 201, solder ball 205, solder paste 207, pad 209, and printed circuit board 211. The x-ray device 102 is configured to capture a respective 2-d gray scale image at each of planes 203-1, 203-2, and 203-N, where N is the total number of planes.” Note that since Kelly doesn’t specify where the planes are have to be located, one could easily have the third image stack be from a plane halfway between the plane of the first image stack and second image stack.) Regarding Claim 6, Kelly in view of Zhang and Wang disclose The method of claim 1, wherein combining includes mapping the first greyscale image stack to a first color channel, mapping the second greyscale image stack to a second color channel, and mapping the third greyscale image stack to a third color channel, wherein the color image includes colors of the first color channel, the second color channel, and the third color channel. (See Kelly Col 3 Lines 55-60, “The color image generator 110 generates a color image using the images 320-1, 320-2 and 320-N. In particular, the color image generator 110 is configured to use one of each of the images 320-1, 320-2, and 320-N as an input into a respective color channel of a color image 324, shown in FIG. 3.”) Regarding Claim 13, Kelly in view of Zhang and Wang disclose The method of claim 1, further comprising determining an anomaly detection threshold for the set of color patches using the neural network model. (See Zhang Page 5 Right Column “B. Results” Paragraph 1, “Results of two models show in Table V and Table VI. From the tables, there is a tradeoff between Recall and F P R. As threshold increases, the model can achieve higher Recall, and F P R increases respectively. For different models, the threshold has different impacts. For example, for 3D CNN model, when threshold is 0.1, on validation dataset, its Recall = 0.9795, while for LSTM based model, Recall = 0.9826.” Also see Zhang Page 6 Tables V-VII showing different thresholds used. Note that this can be considered as “determining an anomaly detection threshold for the set of color patches using the neural network model” as the user can determine the threshold based on the neural network model. Also see Wang Col 4 Lines 21-29, “Furthermore, the method includes determining that the solder joint of the electronic device has no defect only if none of initial prediction probabilities of the solder joint image corresponding to all defect labels is greater than a threshold value. Moreover, the method includes determining that the solder joint of the electronic device has a defect if one of initial prediction probabilities of the solder joint image corresponding to all defect labels is greater than the threshold value.” Also see Wang Col 14 Lines 61-62, “The threshold value can be empirically obtained.” The motivation to combine would have been similar to that of Claim 1 rejection motivation.) Regarding Claim 14, Kelly in view of Zhang and Wang disclose A method comprising: discretizing a three-dimensional (3D) x-ray model of a plurality of ball grid array (BGA) solder balls to generate a plurality of image stacks, wherein each image stack is a two-dimensional (2D) x-ray image; (See Kelly Col 1 Lines 14-23, “One conventional way of inspecting PCBAs involves the use of automated x-ray inspection equipment (AXI). The AXI typically captures 2-dimensional (2-d) gray scale images of interconnects which can then be analyzed using a simple linear regression model. If a defect is identified using the linear regression model and the 2-d images, then a second, offline inspection is performed using a 3-dimensional (3-d) x-ray image to verify the presence of a defect. During the second, offline inspection the PCBA is rotated to obtain the 3-d x-ray image.” Also see Zhang Page 3 Fig. 2 “An example of 3D X-ray imaging with four slices.”) converting each of the plurality of image stacks into a respective greyscale image of each stack; (See Kelly Col 1 Lines 16-17, “The AXI typically captures 2-dimensional (2-d) gray scale images . . .”) combining each of the greyscale image stacks into a color image, wherein the color image is a 2D image; (See Kelly Abstract, “The method further comprises converting the plurality of 2-dimensional gray scale images into a color image. Each of the plurality of 2-dimensional gray scale images corresponds to and is used as input for a respective color channel of the color image.”) identifying a plurality of image patches within the color image, wherein the plurality of image patches includes visual representations of the plurality of BGA solder balls; (See Kelly Col 4 Lines 9-18, “The color image analyzer 108 is configured to identify defects, such as example defect 326, based on analysis of the color image 324. In particular, the color image analyzer 108 implements machine learning techniques to identify defects in the color image 324. The machine learning techniques can include algorithms or models that are generated by performing supervised, unsupervised, or semi-supervised learning on a dataset, and subsequently applying the generated algorithm or model to identify defects, such as the head-in-pillow defect discussed above.” See Zhang Page 2 Left Column Paragraph 2, “Recently, deep learning such as convolutional neural networks (CNNs) have shown outstanding performance on image based tasks such as image classification and object detection. . .They incorporate the whole image as well as region of interest patches to classify whether a joint is qualified or unqualified. . . It requires manual labeling not only for the whole image, but also for each image, four to five patches are required manual labeling.”) providing the plurality of image patches to a trained neural network model; executing the trained neural network model to determine whether each image patch of the plurality of image patches includes at least one anomalous solder ball; and (See Kelly Col 4 Lines 9-18, “The color image analyzer 108 is configured to identify defects, such as example defect 326, based on analysis of the color image 324. In particular, the color image analyzer 108 implements machine learning techniques to identify defects in the color image 324. The machine learning techniques can include algorithms or models that are generated by performing supervised, unsupervised, or semi-supervised learning on a dataset, and subsequently applying the generated algorithm or model to identify defects, such as the head-in-pillow defect discussed above.” Note that for a trained neural network model, it would be obvious to actually use it and provide image patches (taught by Zhang) to determine whether each image patch of the plurality of image patches includes at least one anomalous solder ball.) highlighting, in the 3D x-ray model, at least one anomalous solder ball in at least one image patch to have a possible defect determined by the trained neural network model. (See Kelly Col 4 Lines 9-18, “The color image analyzer 108 is configured to identify defects, such as example defect 326, based on analysis of the color image 324. In particular, the color image analyzer 108 implements machine learning techniques to identify defects in the color image 324. The machine learning techniques can include algorithms or models that are generated by performing supervised, unsupervised, or semi-supervised learning on a dataset, and subsequently applying the generated algorithm or model to identify defects, such as the head-in-pillow defect discussed above.” In this case, identify defect corresponds with highlighting at least one anomalous solder ball in at least one image patch as a defect is determined and selected (thus corresponding to highlighting). The motivation to combine would have been similar to that of Claim 1 rejection motivation.) Regarding Claim 15, Kelly in view of Zhang and Wang disclose The method of claim 14, further comprising determining an anomaly score for the at least one image patch having the at least one anomalous solder ball. (See Zhang Page 5 Right Column “B. Results” Paragraph 1, and Page 6 Tables V-VII teaching different threshold values for determining an anomalous solder ball. Also see Wang Col 4 Lines 21-29 teaching a threshold value. Since there is a threshold, that implies some sort of score, and thus “anomaly score”. The motivation to combine would have been similar to that of Claim 1 rejection motivation.) Regarding Claim 16, Kelly in view of Zhang and Wang disclose The method of claim 15, further comprising determining whether the anomaly score for the at least one image patch is greater than an anomaly detection threshold. (See Wang Col 15 Lines 11-12, “When at least one initial prediction probability of one type of defect label is greater than the preset threshold . . .” The motivation to combine would have been similar to that of Claim 1 rejection motivation.) Regarding Claim 17, Kelly in view of Zhang and Wang disclose The method of claim 16, wherein, if the anomaly score for the at least one image patch is greater than the anomaly detection threshold, indicating that the at least one image patch includes the at least one anomalous solder ball. (See Wang Col 4 Lines 21-29, “Furthermore, the method includes determining that the solder joint of the electronic device has no defect only if none of initial prediction probabilities of the solder joint image corresponding to all defect labels is greater than a threshold value. Moreover, the method includes determining that the solder joint of the electronic device has a defect if one of initial prediction probabilities of the solder joint image corresponding to all defect labels is greater than the threshold value.” The motivation to combine would have been similar to that of Claim 1 rejection motivation.) Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Kelly in view of Zhang and Wang and in further view of Lichtenwalter et al. (“Detect industrial defects at low latency with computer vision at the edge with Amazon SageMaker Edge”) (Hereinafter referred to as Lichtenwalter) Regarding Claim 12, Kelly in view of Zhang and Wang fail to explicitly disclose The method of claim 1, further comprising augmenting the at least one first color patch or the at least one second color patch to modify a feature of the respective color patch. Lichtenwalter teaches The method of claim 1, further comprising augmenting the at least one first color patch or the at least one second color patch to modify a feature of the respective color patch. (Page 2, “Methodology” Paragraph 2, “The images captured for a supervised ML task must include examples of both defective and non-defective products. The defective images are further up sampled by data augmentation techniques to reduce the class imbalance.” Note that augmenting training data is a common practice within the field of machine learning, and in combination with Kelly in view of Zhang and Wang already teaching the at least one first color patch or the at least one second color patch, the above limitations are taught.) It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Kelly in view of Zhang and Wang with Lichtenwalter to include augmenting the image patches. The motivation to combine Kelly in view of Zhang and Wang with Lichtenwalter would have been obvious as all the arts are within the same field of detecting defects using neural networks (See Lichtenwalter Pages 4-5). The benefit of augmenting the image patches and it can be increase the amount of training and thus generally improve the accuracy of the model, see Lichtenwalter Page 2, “Methodology” Paragraph 2 describing how data augmentation can be used to up sample. Allowable Subject Matter Claims 7-11 and 18 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including 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: Regarding Claim 7, the cited prior art does not disclose or render obvious the combination of elements cited in the claims as a whole. Specifically, the cited prior art fails to disclose or render obvious the limitations: masking the first greyscale image stack to mark the at least one solder ball as anomalous, masking the second greyscale image stack to mark the at least one solder ball as anomalous, and masking the third greyscale image stack to mark the at least one solder ball is anomalous. Thus, Claim 7 contains allowable subject matter. Regarding Claims 8-9, Claims 8-9 are dependent upon the base Claim of 7 and therefore also contain allowable subject matter. Regarding Claim 10, the cited prior art does not disclose or render obvious the combination of elements cited in the claims as a whole. Specifically, the cited prior art fails to disclose or render obvious the limitations: locating the plurality of BGA solder balls within the 3D x-ray model to generate a BGA masked image. Thus, Claim 10 contains allowable subject matter. Regarding Claim 11, Claim 11 is dependent upon Claim 10 and therefore also contain allowable subject matter. Regarding Claim 18, the cited prior art does not disclose or render obvious the combination of elements cited in the claims as a whole. Specifically, the cited prior art fails to disclose or render obvious the limitations: including the anomaly score for the at least one image patch as highlighted in the 3D x-ray model. Thus, Claim 18 contains allowable subject matter. Claims 19-20 allowed. The following is an examiner’s statement of reasons for allowance: Claim 19 recites the limitations of combining a plurality of greyscale image stacks of a first three-dimensional (3D) x-ray model of a plurality of ball grid array (BGA) solder balls into a color image, wherein the color image is two-dimensional (2D); determining at least one solder ball as anomalous within one of the plurality of greyscale image stacks; identifying at least one first smaller color image patch within the color image as including at least one anomalous solder ball and at least one second smaller color image patch within the color image as including at least one normal solder ball; training a neural network model with a set of color image patches including the at least one first smaller color image patch having the at least one anomalous solder ball and the at least one second smaller color image patch having the at least one normal solder ball, wherein the neural network model is trained to predict an anomaly within the first 3D x-ray model of the plurality of BGA solder balls; providing a plurality of image patches from a second 3D x-ray model to the trained neural network model; executing the trained neural network model to determine whether each image patch of the plurality of image patches includes at least one anomalous solder ball; and highlighting, in the second 3D x-ray model, the at least one anomalous solder ball in at least one image patch to have a possible defect determined by the trained neural network model. Specifically, the limitations of providing a plurality of image patches from a second 3D x-ray model to the trained neural network model and highlighting, in the second 3D x-ray model, the at least one anomalous solder ball in at least one image patch to have a possible defect determined by the trained neural network model is what renders the claim and its dependent claims, novel and non-obvious as the prior art of record does not disclose or render obvious, the combination of elements recited in the claims as a whole. 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 Any inquiry concerning this communication or earlier communications from the examiner should be directed to THANG G HUYNH whose telephone number is (571)272-5432. The examiner can normally be reached Mon-Thu 7:30am-4:30pm EST | Fri 7:30am-11:30am 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, Kee Tung can be reached at (571)272-7794. 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. /T.G.H./Examiner, Art Unit 2611 /KEE M TUNG/Supervisory Patent Examiner, Art Unit 2611
Read full office action

Prosecution Timeline

Mar 28, 2024
Application Filed
Jan 02, 2026
Non-Final Rejection — §103 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12597100
DEEP IMAGE DELIGHTING
2y 5m to grant Granted Apr 07, 2026
Patent 12586309
MACHINE-LEARNING METHOD ON VECTORIZED THREE-DIMENSIONAL MODEL AND LEARNING SYSTEM THEREOF
2y 5m to grant Granted Mar 24, 2026
Patent 12581083
METHOD, DEVICE, AND COMPUTER PROGRAM PRODUCT FOR COMPRESSING TWO-DIMENSIONAL IMAGE
2y 5m to grant Granted Mar 17, 2026
Patent 12560450
METHOD AND SERVER FOR GENERATING SPATIAL MAP
2y 5m to grant Granted Feb 24, 2026
Patent 12554815
DEVICES, METHODS, AND GRAPHICAL USER INTERFACES FOR AUTHORIZING A SECURE OPERATION
2y 5m to grant Granted Feb 17, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

AI Strategy Recommendation

Get an AI-powered prosecution strategy using examiner precedents, rejection analysis, and claim mapping.
Powered by AI — typically takes 5-10 seconds

Prosecution Projections

1-2
Expected OA Rounds
76%
Grant Probability
99%
With Interview (+50.0%)
2y 4m
Median Time to Grant
Low
PTA Risk
Based on 25 resolved cases by this examiner. Grant probability derived from career allow 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