Prosecution Insights
Last updated: November 18, 2026
Application No. 18/602,631

OBJECT DETECTION IN DOCUMENTS USING NEURAL NETWORKS

Final Rejection §103
Filed
Mar 12, 2024
Examiner
SALEH, ZAID MUHAMMAD
Art Unit
2668
Tech Center
2600 — Communications
Assignee
Abbyy Development Inc.
OA Round
2 (Final)
68%
Grant Probability
Favorable
3-4
OA Rounds
99%
With Interview

Examiner Intelligence

Grants 68% — above average
68%
Career Allowance Rate
44 granted / 65 resolved
+5.7% vs TC avg
Strong +43% interview lift
Without
With
+42.9%
Interview Lift
resolved cases with interview
Typical timeline
Avg Prosecution
24 currently pending
Career history
103
Total Applications
across all art units

Office Action

§103
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 . Status of Claims Claims 1 – 12 and 14 – 20 remain pending. Claims 1, 4, 6, 7, 12, 15 – 17, 19, and 20 are Amended Claim 13 have been canceled. Response to Arguments Applicant's arguments filed July 06, 2026 with respect to claims 1 – 12 and 14 – 20 have been considered but are moot because the new grounds of rejection do not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. Claim Rejections - 35 USC § 103 In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status. The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. Claims 1, 6, 7, 20 and 21 are rejected under 35 U.S.C 103 as being unpatentable over Sarkar et al. US Patent Publication No. US-20190294661-A1 (hereinafter Sarkar) in view of Ye US Patent Application Publication No. US-20210295114-A1 (hereinafter Ye) and Zhang Patent Application Publication No. CN-111898547-A (hereinafter Zhang) and further in view of Xu Patent Application Publication No. CN-115631493-A (hereinafter Xu). Regarding claim 1, Sarkar discloses about a method comprising: processing, using a machine learning model (MLM), a representation of an image of at least a portion of a document to generate a plurality of pixel-level maps (PLMs) (Sarkar in [0051] discloses, “the form conversion system 106 generates one or more semantic segmentation maps classifying each pixel as a particular element type based on the determined semantic characteristics of that pixel”. Furthermore, Sarkar discloses in [0058], “the form conversion neural network 408 determines element type probabilities for that pixel. For example, the form conversion neural network 408 may determine that a pixel of the digitized paper form 402 has a probability of 0.5 that it is part of a widget, a probability of 0.3 that it is part of a text run, and a probability of 0.2 that it is part of an image”), and a plurality of classification NNs, each classification NN of the plurality of classification NNs processing the feature tensor and generating one or more PLMs of the plurality of PLMs (Sarkar in [0025] discloses about neural network decoders processing feature map (feature tensor), “a neural network encoder that processes the digitized paper form and outputs a feature map, which is then fed to several neural network decoders that further process the feature map”. Additionally, in [0027] Sarkar discloses each semantic decoder classifies pixel (classification NN), “the form conversion system can train the neural network semantic decoders ... classify each pixel (i.e., determine to which element type each pixel belongs) based on the determined semantic characteristics”), and generating, using the plurality of PLMs, an object-level map identifying placement of one or more objects in the document (Sarker in [0057] discloses, “the semantic segmentation map 410 provides an indication of the element type to which the pixels of the digitized paper form 402 correspond”. Additionally, in [0051] Sarkar discloses, “the form conversion system 106 generates one or more semantic segmentation maps classifying each pixel as a particular element type based on the determined semantic characteristics of that pixel and generates the digital fillable form based on those semantic segmentation maps”). Sarkar doesn’t disclose the following limitation as further recited in the claim. Ye discloses the MLM comprises: a backbone neural network (NN) processing the representation of the image and generating a feature tensor representative of the image (Ye discloses about backbone neural network generating a feature tensor in [0050] and [0051], “the image text extraction model is a multi-class deep neural network, and includes a backbone network” and “After the input image passes through the backbone network, one or more feature tensors may be output”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Ye into the system of Sarkar because it would improve the models ability to understand documents content beyond pixels. Sarkar and Ye in the combination don’t disclose the following limitation as further recited in the claim. Zhang discloses about at least a subset of the plurality of classification NNs is trained together with the backbone NN (Zhang in [0133] discloses, “the face recognition model includes a backbone network and a plurality of classification networks”. Furthermore, Zhang in [0089] discloses, “calculating the classification loss function value of each classification data set to obtain a plurality of classification loss function values”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Zhang into the system of Sarkar in view of Ye because it would allow the system to be more accurate in detecting diverse document objects. Sarkar, Ye and Zhang in the combination don’t disclose the following limitation as further recited in the claim. Xu discloses wherein the plurality of PLMs comprises: one or more first PLMs, each PLM of the one or more first PLMs comprising, for an individual pixel of the image, a respective likelihood of the individual pixel being associated with a respective object type of one or more object types, and a second PLM comprising, for the individual pixel of the image, a first likelihood of the individual pixel belonging to a same-type object as an adjacent pixel of the image (Xu in [Page – 2, Paragraph – 5] discloses, “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Xu into the system of Sarkar in view of Ye and Zhang because it would allow the system to perform more accurate and efficient reconstruction of document objects from pixel level prediction. Summary of Citations (Xu) [Page – 2, Paragraph – 5]; “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”. Summary of Citations (Zhang) Paragraph [0089]; “And calculating the feature vector loss function value of each feature set to obtain a plurality of feature vector loss function values, and calculating the classification loss function value of each classification data set to obtain a plurality of classification loss function values”. Paragraph [0133]; “the face recognition model includes a backbone network and a plurality of classification networks”. Summary of Citations (Ye) Paragraph [0050]; “the image text extraction model is a multi-class deep neural network, and includes a backbone network”. Paragraph [0051]; “After the input image passes through the backbone network, one or more feature tensors may be output”. Summary of Citations (Sarkar) Paragraph [0025]; “a neural network encoder that processes the digitized paper form and outputs a feature map, which is then fed to several neural network decoders that further process the feature map”. Paragraph [0027]; “the form conversion system can train the neural network semantic decoders to determine the semantic characteristics of each pixel of the digitized paper form and then classify each pixel (i.e., determine to which element type each pixel belongs) based on the determined semantic characteristics”. Paragraph [0051]; “the form conversion system 106 generates one or more semantic segmentation maps classifying each pixel as a particular element type based on the determined semantic characteristics of that pixel and generates the digital fillable form based on those semantic segmentation maps”. Paragraph [0057]; “the semantic segmentation map 410 provides an indication of the element type to which the pixels of the digitized paper form 402 correspond”. Paragraph [0058]; “the form conversion neural network 408 determines element type probabilities for that pixel. For example, the form conversion neural network 408 may determine that a pixel of the digitized paper form 402 has a probability of 0.5 that it is part of a widget, a probability of 0.3 that it is part of a text run, and a probability of 0.2 that it is part of an image”. Regarding claim 6, the combination of Sarkar in view of Ye and Zhang and further in view Xu as a whole teaches claim 1, and Sarkar teaches claim 6 for the same grounds of rejection from the Non-Final Office Action of 04/08/2026. Regarding claim 7, Xu in the combination discloses the method of claim 1, wherein generating the object-level map comprises: associating the individual pixel with the adjacent pixel based at least on (i) the first likelihood and (ii) a second likelihood of the adjacent pixel belonging to the same-type object as the individual pixel of the image (Xu in [Page – 2, Paragraph – 5] discloses, “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”). Regarding claim 20, apparatus claim 20 corresponds to method claim 1. Therefore, the rejection analysis and motivation to combine of claim 1 is applicable to claim 20. Regarding claim 21, Xu in the combination discloses the system of claim 20, wherein generating the object-level map comprises: associating the individual pixel with the adjacent pixel based at least on (i) the first likelihood and (ii) a second likelihood of the adjacent pixel belonging to the same- type object as the individual pixel of the image (Xu in [Page – 2, Paragraph – 5] discloses, “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”). Summary of Citations (Xu) [Page – 2, Paragraph – 5]; “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”. Claim 2 is rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Ye, Zhang and Xu and further in view of Torres US Patent Application Publication No. US-20210034856-A1 (hereinafter Torres). Regarding claims 2, the combination of Sarkar in view of Ye, Zhang and Xu as a whole teaches claim 1 but fails to teach the further limitations as recited in claims 2. Torres teaches claim 2 for the same grounds of rejection and motivation established in the Non-Final Office Action of 04/08/2026. Claims 3 and 4 are rejected under 35 U.S.C 103 as being unpatentable over Sarker in view of Ye, Zhang, Xu and Torres and further in view of Sarkar US Patent Publication No. US-20190050640-A1 (hereinafter Sarkar640). Regarding claims 3 and 4, the combination of Sarkar in view of Ye, Zhang, Xu and Torres as a whole teaches claim 1 but fails to teach the further limitations as recited in claims 3 and 4. Sarkar640 teaches claims 3 and 4 for the same grounds of rejection and motivation established in the Non-Final Office Action of 04/08/2026. Claim 5 is rejected under 35 U.S.C 103 as being unpatentable over Sarker in view of Ye, Zhang, Xu, Torres and Sarkar640 and further in view of Jin Patent Application Publication No. CN-116524503-A (hereinafter Jin). Regarding claims 5, the combination of Sarkar in view of Ye, Zhang, Xu, Torres and Sarkar640 as a whole teaches claim 1 but fails to teach the further limitations as recited in claim 5. Jin teaches claim 5 for the same grounds of rejection and motivation established in the Non-Final Office Action of 04/08/2026. Claims 8, 9, 10 and 11 are rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Ye, Zhang and Xu and further in view of Friedman US Patent Publication No. US-9530184-B1 (hereinafter Friedman). Regarding claims 8 – 11, the combination of Sarkar in view of Ye, Zhang and Xu as a whole teaches claim 1 but fails to teach the further limitations as recited in claims 2. Friedman teaches claims 8 – 11 for the same grounds of rejection and motivation established in the Non-Final Office Action of 04/08/2026. Claims 12 and 18 are rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Zagaynov US Patent Publication No. US-20220366179-A1 (hereinafter Zagaynov) and Deng Patent Application Publication No. JP-2021125007-A and further in view of Rangarajan US Patent Publication No. US-12300008-B1 (hereinafter Rangarajan). Regarding claim 12, Sarkar discloses method comprising: processing, using a machine learning model (MLM), a first patch of the plurality of patches to generate a first plurality of pixel-level maps (PLMs), each PLM of the first plurality of PLMs characterizing associations of pixels of the first patch with a respective object type of a plurality of object types; processing, using the MLM, a second patch of the plurality of patches to generate a second plurality of PLMs, each PLM of the second plurality of PLMs characterizing associations of pixels of the second patch with the respective object type of the plurality of object types (Sarkar in [0051] discloses, “the form conversion system 106 generates one or more semantic segmentation maps classifying each pixel as a particular element type based on the determined semantic characteristics of that pixel”. Furthermore, Sarkar discloses in [0058], “the form conversion neural network 408 determines element type probabilities for that pixel. For example, the form conversion neural network 408 may determine that a pixel of the digitized paper form 402 has a probability of 0.5 that it is part of a widget, a probability of 0.3 that it is part of a text run, and a probability of 0.2 that it is part of an image”). Sarkar doesn’t disclose the following limitation as further recited in the claim. Zagaynov discloses identifying a size of a text depicted in an image of a document (Zagaynov in [0091] discloses, “the one or more second feature vectors may include text size feature vectors which may be representative of the size of the text in the image”); representing, based at least on the identified size of the text, the image via a plurality of patches of a target size (Zagaynov in [0046] discloses, “patches of various sizes may be resized to the same common size, e.g., 128×128 size, or 256×256 size, or any other pre-set common size”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Zagaynov into the system of Sarkar because it would allow the system to detect document object reliably even when the document have different text sizes or handwriting size. Sarkar and Zagaynov in the combination don’t disclose the following limitation as further recited in the claim. Deng discloses aggregating the first plurality of PLMs and the second plurality of PLMs to obtain at least one aggregated PLM, wherein the aggregated PLM comprises an aggregated value associated with an individual pixel of an overlapping region of the first patch and the second patch, the aggregated value obtained using (i) one or more first values associated with the individual pixel and the first plurality of PLMs and (ii) one or more second values associated with the individual pixel and the second plurality of PLMs (Deng in [0048] and [0049] discloses about aggregated value associated with an individual pixel of an overlapping region, “These first a modal masks 401a and 401b represent the probability values that each pixel belongs to a specific object. More specifically, when two objects A and B partially overlap, in the a modal mask generated for object A” and “the mask aggregation unit according to the present invention combines the first a modal masks 401a and 401b with each other and adds the probabilities for each pixel to combine the first a modal masks 401a and 401b, respectively ... the boundary (coordinates, etc.) of the overlapping area where each object overlaps can be identified with high accuracy, and the overlapping mask described later can be generated more easily”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Deng into the system of Sarkar in view of Zagaynov because it would allow the system to build a more reliable object level map of the document. Sarkar, Zagaynov and Deng in the combination don’t disclose the following limitation as further recited in the claim. Rangarajan discloses generating, using the at least one aggregated PLM, an object-level map identifying location of one or more objects in the document (Rangarajan in [Column – 8, Line 14 – 17] discloses, “The locations of the other sensitive data elements was identified by applying the region detection model 155 and element detection model 165, and combining the results of both models”. Furthermore, Rangarajan in [Column – 14, Line 14 – 18] discloses, “For each identified text string, the parser may return its location. The location of the string returned by the parser may be mapped to pixel coordinates in a page of the document image file”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Rangarajan into the system of Sarkar in view of Zagaynov and Deng because it would allow the system to identify where different objects are located in the document. Summary of Citations (Zagaynov) Paragraph [0046]; “patches of various sizes may be resized to the same common size, e.g., 128×128 size, or 256×256 size, or any other pre-set common size”. Paragraph [0091]; “the one or more second feature vectors may include text size feature vectors which may be representative of the size of the text in the image”. Summary of Citations (Rangarajan) [Column – 8, Line 14 – 17]; “The locations of the other sensitive data elements was identified by applying the region detection model 155 and element detection model 165, and combining the results of both models”. [Column – 14, Line 14 – 18]; “For each identified text string, the parser may return its location. The location of the string returned by the parser may be mapped to pixel coordinates in a page of the document image file”. Summary of Citations (Deng) Paragraph [0048]; “These first a modal masks 401a and 401b represent the probability values that each pixel belongs to a specific object. More specifically, when two objects A and B partially overlap, in the a modal mask generated for object A” Paragraph [0049]; “the mask aggregation unit according to the present invention combines the first a modal masks 401a and 401b with each other and adds the probabilities for each pixel to combine the first a modal masks 401a and 401b, respectively ... the boundary (coordinates, etc.) of the overlapping area where each object overlaps can be identified with high accuracy, and the overlapping mask described later can be generated more easily”. Regarding claim 18, the combination of Sarkar in view of Zagaynov, Rangarajan and Deng as a whole teaches claim 1, and Sarkar teaches claim 18 for the same grounds of rejection from the Non-Final Office Action of 04/08/2026. Claim 14 is rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Zagaynov, Rangarajan, Deng, Friedman, and further in view of Raedt WO-2021053135-A1 (hereinafter Raedt). Regarding claims 14, the combination of Sarkar in view of Zagaynov, Rangarajan and Deng as a whole teaches claim 14 but fails to teach the further limitations as recited in claims 14. Friedman and Raedt teaches claim 14 for the same grounds of rejection and motivation established in the Non-Final Office Action of 04/08/2026. Claims 15 and 19 are rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Zagaynov, Rangarajan, Deng, and further in view of Xu Regarding claim 15, Sarkar in the combination discloses the method of claim 12. Sarkar, Zagaynov, Deng and Rangarajan in the combination don’t disclose the following limitation as further recited in the claim. Xu discloses generating the object-level map comprises: associating an individual pixel of the first patch with an adjacent pixel based at least on (i) a first likelihood of the individual pixel belonging to a same-type object as the adjacent pixel and (ii) a second likelihood of the adjacent pixel belonging to the same- type object as the individual pixel (Xu in [Page – 2, Paragraph – 5] discloses, “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Xu into the system of Sarkar in view of Zagaynov, Rangarajan and Deng because it would allow the system to perform more accurate and efficient reconstruction of document objects from pixel level prediction. Summary of Citations (Xu) [Page – 2, Paragraph – 5]; “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”. Regarding claim 19, Sarkar in the combination discloses the method of claim 12. Sarkar, Zagaynov, Deng and Rangarajan in the combination don’t disclose the following limitation as further recited in the claim. Xu discloses the first plurality of PLMs comprises a PLM that includes, for the individual pixel, a first likelihood of neighboring the individual pixels of the first patch belonging to a same-type object as an adjacent pixel (Xu in [Page – 2, Paragraph – 5] discloses, “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”). Summary of Citations (Xu) [Page – 2, Paragraph – 5]; “predicting the text image to be tested by the target prediction model, obtaining the first probability of each pixel point in the text image to be tested belonging to the text area, and the second probability of each pixel point and the adjacent pixel point belong to the same type, belonging to the same type”. Claim 16 is rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Zagaynov, Rangarajan, Deng, and further in view of Sarkar US Patent Publication No. US-20190050640-A1 (hereinafter Sarkar640). Regarding claim 16, Sarkar640 discloses the method of claim 12, Sarkar, Zagaynov, Deng and Rangarajan in the combination don’t disclose the following limitation as further recited in the claim. Sarkar640 discloses the overlapping region is between the first patch, the second patch, a third patch, and a fourth patch (Sarakar640 in [0099] discloses, “generating a plurality of tiles from said image, each of said tiles comprising a subset of pixels from said high-resolution image”. Sarkar640 in [0031] discloses, “the tiles may overlap with one another”), and wherein the aggregated value is further obtained using (i) one or more third values associated with the individual pixel and the third plurality of PLMs generated for the third patch and (ii) one or more fourth values associated with the individual pixel and the fourth plurality of PLMs generated for the fourth patch (Sarkar640 in [0100] discloses, “processing said second feature map to generate class predictions for each pixel in said tile and, aggregating each of said respective predictions for each pixel of said high-resolution image to generate a global feature map for said document”). Summary of Citations (Sarkar640) Paragraph [0031]; “the tiles may overlap with one another”. Paragraph [0099]; “generating a plurality of tiles from said image, each of said tiles comprising a subset of pixels from said high-resolution image”. Paragraph [0100]; “processing said second feature map to generate class predictions for each pixel in said tile and, aggregating each of said respective predictions for each pixel of said high-resolution image to generate a global feature map for said document”. Claim 17 is rejected under 35 U.S.C 103 as being unpatentable over Sarkar in view of Zagaynov, Rangarajan, Deng, and further in view of Jin Patent Application Publication No. CN-116524503-A (hereinafter Jin). Regarding claim 17, Sarkar in the combination the method of claim 12. Sarkar, Zagaynov, Deng and Rangarajan in the combination don’t disclose the following limitation as further recited in the claim. Jin discloses aggregating a first value of the one or more first values and a second value of the one or more second values comprises at least one of: selecting a maximum value of the first value and the second value as the aggregated value; selecting a minimum value of the first value and the second value as the aggregated value; or selecting a weighted combination of the first value and the second value as the aggregated value, wherein weights in the weighted combination are determined based on a location of the individual pixel within the overlapping portion (Jin in [0016 – 0017] discloses, “aggregating the first text character candidate boxes with partial overlap into a candidate box set; obtaining a confidence score of each first text character candidate frame in the candidate frame set, and taking the first text character candidate frame with the highest confidence score as a current candidate frame; the confidence score is the probability that the pixel image contained in the first text character candidate box is a complete text character”). It would have been obvious to one of ordinary skill in art before the effective filling date of the claimed invention to integrate the technique of Jin into the system of Sarkar in view of Zagaynov, Rangarajan and Deng because using the maximum value keeps the strongest probability among the overlapping regions. Summary of Citations (Jin) Paragraph [0016 – 0017]; “aggregating the first text character candidate boxes with partial overlap into a candidate box set; obtaining a confidence score of each first text character candidate frame in the candidate frame set, and taking the first text character candidate frame with the highest confidence score as a current candidate frame; the confidence score is the probability that the pixel image contained in the first text character candidate box is a complete text character”. Conclusion Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a). A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) 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. Contact Any inquiry concerning this communication or earlier communications from the examiner should be directed to ZAID MUHAMMAD SALEH whose telephone number is (703)756-1684. The examiner can normally be reached M-F 8 am - 5 pm ET. 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 on (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. /ZAID MUHAMMAD SALEH/ Examiner, Art Unit 2668 9/13/2026 /VU LE/Supervisory Patent Examiner, Art Unit 2668
Read full office action

Prosecution Timeline

Mar 12, 2024
Application Filed
Apr 08, 2026
Non-Final Rejection mailed — §103
Jul 06, 2026
Response Filed
Jul 10, 2026
Examiner Interview Summary
Sep 17, 2026
Final Rejection mailed — §103 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12770245
AUTOMATED VISUAL-INSPECTION SYSTEM
3y 11m to grant Granted Oct 20, 2026
Patent 12755570
UNIFIED FRAMEWORK FOR ANALYSIS AND RECOGNITION OF IDENTITY DOCUMENTS
3y 11m to grant Granted Oct 06, 2026
Patent 12749221
EXTRINSIC CAMERA CALIBRATION USING CALIBRATION OBJECT
3y 11m to grant Granted Sep 29, 2026
Patent 12725290
METHOD AND APPARATUS FOR ESTIMATING A BODY PART POSITION OF A PERSON
2y 7m to grant Granted Sep 01, 2026
Patent 12725430
CORRECTING AN ALIGNMENT OF POSITIONS OF POINTS AFFILIATED WITH AN OBJECT, IN IMAGES OF A LOCATION, THAT HAS A LINEAR FEATURE OR A PLANAR FEATURE
2y 12m to grant Granted Sep 01, 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
68%
Grant Probability
99%
With Interview (+42.9%)
Moderate
PTA Risk
Based on 65 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