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
This action is in reply to the communications filed on January 23, 2026, and March 17, 2026. The Applicants’ Amendment and Request for Reconsideration has been received and entered.
Claims 1, 6-7, 9, 14-15, and 21-30 are currently pending and have been examined. Claims 1, 9, and 22 have been amended.
The previous objection to claim 9 has been withdrawn.
The previous rejection of claims 1, 6-7, 9, 14-15, 21-22, 26, and 29-30 under 35 USC 112(a) has been withdrawn.
The previous rejection of claims 1, 6-7, 9, 14-15, and 21-30 under 35 USC 112(b) has been withdrawn.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on January 23, 2026, has been entered.
Response to Arguments
The previous objection to claim 9 has been withdrawn in view of the amendments to claim 9, including the amendment to the status identifier.
The previous rejection of claims 1, 6-7, 9, 14-15, 21-22, 26, and 29-30 under 35 USC 112(a) has been withdrawn in view of Applicants’ amendments and arguments.
The previous rejection of claims 1, 6-7, 9, 14-15, and 21-30 under 35 USC 112(b) has been withdrawn in view of Applicants’ amendments and arguments.
Applicants’ arguments regarding the rejections of claims 23-25 and 27-28 have been fully considered but they are not persuasive. Per MPEP 2163: "To satisfy the written description requirement, a patent specification must describe the claimed invention in sufficient detail that one skilled in the art can reasonably conclude that the inventor had possession of the claimed invention." Further: "When an explicit limitation in a claim is not present in the written description whose benefit is sought it must be shown that a person of ordinary skill would have understood, at the time the patent application was filed, that the description requires that limitation." Hyatt V. Boone, 146 F.3d 1348.
With this guidance in mind, the Examiner respectfully asserts that, regarding claims 23-25 and 27-28, Applicants have not shown that the description requires these limitations.
With respect to claim 23, Applicants have argued at page 9 of Applicants’ Reply dated January 23, 2026 (hereinafter “Applicants’ Reply”) that the “Specification’s disclosure of product ‘locations,’ bounding boxes, and neural-network-based image models…conveys possession of ROI delineations in the broader genus of localized regions, which in computer vision includes non-rectangular masks as a routine alternative to bounding boxes for representing product locations; a POSITA would understand this as within the disclosed localization framework.” The Examiner respectfully disagrees.
Applicants have referenced paragraphs [0041]-[0042] of the Specification. However, bounding boxes are only referenced in paragraph [0040] of the as-filed and published applications, which is reproduced herein:
[0040] The method 500 may further include, at step 504, identifying at least one product in the obtained at least one user content image using at least one artificial intelligence model. The location and the identity of the products are identified in the UGC image. The locations of the products are within bounding box co-ordinates defined in the UGC image. In an embodiment, the bounding box may be drawn by the user of the content management platform 106 on the UGC image. The bounding box allows the artificial intelligence model to narrow the scope of search for a product in the UGC image. The artificial intelligence model may be already trained using product catalogs specific to the user of the content management platform 106. The artificial intelligence model is trained using training data. The training data includes images of products in a plurality of sample UGCs or a product catalog as an input to the artificial intelligence model and identity information of the products in the product catalog as an output of the artificial intelligence model. The training data is obtained from third party service provider websites 112 such as data crowdsourcing service providers, like, FigureEight, RapidWorkers, SamaSource, etc.
Further, the only mention of computer vision is in paragraph [0033] of the as-filed and published applications, which is reproduced herein:
[0033] The UGC may be collected using automated search algorithms to scrape photos, videos, and other content from a variety of social media platforms, including but not limited to Facebook®, Instagram®, and Twitter®, according to the search criteria set by content parameters. The appropriate UGC may be selected for use in the media album. The UGC may be selected either manually or automatically, for inclusion in visual displays of the media album. In some example embodiments, a set of computer software may be used to make this process simple for the clients, if manual selection is used. If automatic, filters which may include computer vision, natural language processing, and other algorithms may be used to narrow the collected UGC into an optimal subset. Once the UGC is selected, permissions for UGC may be obtained. For each piece of UGC to be included in the group (album), permissions may be obtained from the original content creator on the origin social media platform. In some example embodiments, this may be performed through a piece of software which connects the client with the creator on the origin content management platform 106. Further, the UGC may be organized in the media album based on the goal for optimization and the purpose for the media album. That is, the UGC may be optimized for display. In some example embodiments, the content within the media album may be ranked through one of several processes and the resultant ranking may be used to determine content order when displayed. Since in many cases, only a few pieces of content may be displayed at once, ranking is important so that optimal content may be displayed. Content may be grouped based on the number of displayed pieces with each load, and randomized within the group. In some example embodiments, such content may be displayed across a variety of locations, both online and in brick-and-mortar stores, in print, etc..
First, the Examiner respectfully asserts that only bounding boxes are referenced in the application. No other bounding structures or masks are referenced in the application. The Examiner further asserts that the only mention of computer vision is in paragraph [0033] and it is referenced as a filter to narrow content. Computer vision is not mentioned in the context of bounding structures, bounding boxes, or masks and a person having ordinary skill in the art would not have understood that the description requires non-rectangular masks as a routine alternative to bounding boxes for representing product locations merely by mentioning computer vision in an unrelated portion of the specification.
With respect to claim 24, Applicants argue at page 9 of Applicants’ Reply that the “Specification also teaches dynamic ranking and scoring based on historical interaction data and manual confirmations, including multi-armed bandit approaches (Detailed Description, Method 300, [0031]; Method 400, [0036]-[0037]; Method 600, [0046]). In view of the express disclosure of threshold-based subset selection (Detailed Description, [0045]), it would be routine, and is reasonably conveyed by the Specification, to dynamically determine thresholds based on historical user confirmation activity.” The Examiner respectfully disagrees.
The Examiner respectfully asserts that there is no connection between selecting a subset based on a threshold and dynamically determining thresholds based on historical user confirmation activity. In fact, the only mention of a threshold in Applicants’ disclosure is in paragraph [0045] of the as-filed and published applications, which recites that the “subset can be ordered by probability ranks, and may be cut off by some probability threshold.” This is not sufficient to show that the description requires dynamically determining thresholds based on historical user confirmation activity.
With respect to claim 25, Applicants argue at pages 9-10 of Applicants’ Reply that “the Detailed Description discloses matching NTI/NTII in UGC images to a product catalog and associating tags with products….These passages together demonstrate that the system records confirmed tag associations and localized product information (i.e., ROI/location data) as part of its data pipeline and model retraining, and operates over a catalog-backed architecture….A POSITA would understand, in view of the express disclosure of catalog matching and storage components, that the confirmed tag and its associated ROl/location would be stored and linked to the corresponding product entry in the catalog to support subsequent discovery, display, and retraining." The Examiner respectfully disagrees and asserts that, while the disclosure discusses catalog matching and storage components, there is no indication of updating the product catalog and thus this is not sufficient to show that the description requires updating the product catalog to associate the confirmed tag and ROI with the corresponding product entry.
With respect to claim 27, Applicants argue at page 10 of Applicants’ Reply that Figures 8-12 and paragraphs [0046] and [0050]-[0051] support “presenting batches for simultaneous actions which a POSITA would understand to include simultaneous confirmation of multiple ROIs from multiple images." The Examiner respectfully disagrees. The Examiner respectfully asserts that Figures 8-12 are not presenting images for simultaneous confirmation but instead are merely presenting multiple images to the user for display. Similarly, paragraphs [0046] and [0050]-[0051] do not describe simultaneous confirmation either. Paragraph [0046] of the as-filed and published applications, discloses in its entirety:
[0046] In an embodiment, the content management platform 106 may recommend a list of tagged UGC images that may contain the product and display such recommendations in an itemized manner on a user interface of the user device 102. The user of the content management platform 106 may manually select and confirm that the tagged UGC images comprise the product. The artificial intelligence model may be iteratively trained based on an association between each of the plural user content and each of the first one of the plural tags and the generated subset of the tagged UGC images.
Paragraph [0046] does not reference multiple ROIs from multiple images but instead merely references manual confirmation that tagged UGC images comprise the product.
Paragraph [0050] of the as-filed and published applications, discloses in its entirety:
[0050] At step 710, the user of the content management platform 106 may customize the way the reviews constituting the UGC may appear through the design editor of the content management platform 106. The user of the content management platform 106 may also configure review display density. The UGC may appear in the gallery following a randomization algorithm. The review display density may be referred to the number of reviews, or the UGC that can fit into the display area. The visual tiles with UGC images may be mixed with review tiles as exemplarily illustrated in FIG. 8 and FIG. 10. The tile opacity may be adjusted. Based on the output interface, the text in the UGC may be displayed on medium and large tiles.
Paragraph [0050] does not reference multiple ROIs from multiple images but instead merely references that the user may customize the way that the reviews constituting the UGC may appear.
Paragraph [0051] of the as-filed and published applications, discloses in its entirety:
[0051] FIGS. 8-12 illustrate display representations comprising curated UGC, in accordance with different exemplary embodiments. FIG. 8 illustrates an exemplary display representation 800 in which visual tiles, such as visual tile 802, and textual tiles containing a product review, such as tile 804 may be displayed together to provide a gallery of product reviews. FIG. 9 illustrates another exemplary representational layout 900 for presenting visual tile 902 and text tile 904 for product reviews in a side-by-side manner, thus having a different review density. FIGS. 10-11 are other exemplary display representations 1000 and 1100 respectively illustrating different embodiments for presenting curated UGC. FIG. 12 illustrates an interface 1200 generated by the content management platform allowing curation, publication, and analytics on reviews constituting the UGC The interface 1200 allows importing of reviews from other review and rating service providers, manage star ratings of products, and specify filters for review publishing and the like.
Paragraph [0051] does not reference multiple ROIs from multiple images but instead merely references that the display may be curated to represent different tile densities.
Thus, these cited portions are not sufficient to show that the description requires presenting a batch of ROIs from multiple UGC images to the user for simultaneous confirmation.
With respect to claim 28, Applicants argue at page 10 of Applicants’ Reply that “models compute per-tag probabilities for images (Detailed Description [0041]) and provide rich UI displays" and that therefore "presenting explanations or visualizations of those tag probabilities alongside an ROI in the UI is a routine application of the disclosed data and UI capabilities." The Examiner respectfully disagrees
Paragraph [0041] of the as-filed and published applications, discloses in its entirety:
[0041] The artificial intelligence model is trained on samples of UGC images and images in sample product catalogs with labelled product information. The product information may be labelled based on: principal colors or key colors, for example, shown on a clothing item's image; categories of the products, such as, dress, boots, etc., gender of users to whom the products may cater, age of users to whom the products may cater, location of the products in the sample UGC images, visual similarity between different products, etc. The number of labels on a sample image may be greater than 1. The product catalogs are specific to each user of the content management platform 106, that is a customer of a brand, etc. A UGC image is input to at least 5 trained artificial intelligence models. The output of the trained artificial intelligence models is an identified scored list of products and locations of the products relevant to the input UGC image. Each of the trained artificial intelligence models gives output, such as, identifying a product in the UGC with 80% probability as a male, 75% probability that the product is dress, 11% probability the product is a hat, 70% probability that the product is blue in color, and 40% probability that the product is green in color, etc.
Paragraph [0041] does not reference presenting, along with the ROI, an explanation or visualization of the features or tag probabilities that led to the ROI selection but is instead referring to the output of the artificial intelligence models—not of something output on a display. Thus, paragraph [0041] is not sufficient to show that the description requires presenting, along with the ROI, an
explanation or visualization of the features or tag probabilities that led to the ROI selection.
Thus, the rejection of claims 23-25 and 27-28 under 35 USC 112(a) is maintained.
Applicants’ remaining arguments have been fully considered but they have either been addressed above or they are moot in view of the new grounds of rejection.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 23-25 and 27-28 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 23: Claim 23 recites “wherein the ROI is defined by a non-rectangular mask generated by the artificial intelligence model.” Nowhere does Applicants’ originally-filed disclosure recite that the ROI is defined by a non-rectangular mask generated by the artificial intelligence model.
Claim 24: Claim 24 recites “wherein the threshold for tag probability is dynamically determined based on historical user confirmation activity.” Nowhere does Applicants’ originally-filed disclosure recite that the threshold for tag probability is dynamically determined based on historical user confirmation activity.
Claim 25: Claim 25 recites “updating the product catalog to associate the confirmed tag and ROI with the corresponding product entry.” Nowhere does Applicants’ originally-filed disclosure recite updating the product catalog to associate the confirmed tag and ROI with the corresponding product entry.
Claim 27: Claim 27 recites “presenting a batch of ROIs from multiple UGC images to the user for simultaneous confirmation.” Nowhere does Applicants’ originally-filed disclosure recite presenting a batch of ROIs from multiple UGC images to the user for simultaneous confirmation.
Claim 28: Claim 28 recites “presenting, along with the ROI, an explanation or visualization of the features or tag probabilities that led to the ROI selection.” Nowhere does Applicants’ originally-filed disclosure recite presenting, along with the ROI, an explanation or visualization of the features or tag probabilities that led to the ROI selection.
Potentially Allowable Subject Matter
Claims 23-25 and 27-28 would be allowable if rewritten to overcome the rejection(s) under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action and to include all of the limitations of the base claim and any intervening claims.
Allowable Subject Matter
Claims 1, 6-7, 9, 14-15, 21-22, 26, and 29-30 are allowed.
With regard to claim 1, the prior art of record, alone or combined, neither anticipates nor renders obvious, a method for discovering at least one product-associated user-generated-content (UGC) image from plural UGC images so that a user can select a product from the product-associated UGC image, the method comprising: identifying, by one or more processors of a computer system, plural tags associated with UGC for associating with each of plural product-containing UGC images that each contain non-textual image information (NTII) and each relate to textual information (TI), the identifying the plural tags of UGC comprising identifying the plural tags of UGC using the NTII in the product-containing UGC images and an artificial intelligence model; determining, by the one or more processors, the probability that each of the identified plural tags is associated with the products in each of the product-containing UGC images using the NTII in the product-containing UGC images and outputs from each of at least a first artificial intelligence model and a second artificial intelligence model; associating, by the one or more processors, the identified plural tags with the products in each of the product-containing UGC images based upon the identifying and determining steps by: matching, by the one or more processors, the NTII in the product-containing UGC images to a product catalog; selecting, by the one or more processors, a first one of the plural tags associated with a product in a first UGC image; generating, by the one or more processors, at least one subset of the UGC pixels in the first UGC image as a region-of-interest (ROI) defined by bounding-box coordinates automatically generated by the artificial intelligence model, wherein the ROI comprises a pixel- level region cropped around the product whose tag probability for the first one of the plural tags being associated with the product in the first UGC image exceeds a threshold; presenting the ROI for confirmation input from a user that the first one of the plural tags is associated with the product in the first UGC image; obtaining, by the one or more processors, a confirmation input from a user that the first one of the plural tags is associated with the product in the first UGC image; iteratively training, by the one or more processors, the first artificial intelligence model based on an association between each of the plural UGC images and each of the first one of the plural tags, wherein iteratively means training at least two times; iteratively training, by the one or more processors, the second artificial intelligence model based on an association between each of the plural UGC images and each of the first one of the plural tags, wherein iteratively means training at least two times; wherein the at least first artificial intelligence model and the second artificial intelligence model each include at least two of a neural network model, a nearest neighbor model, a k-nearest neighbor clustering model, a singular value decomposition model, a principal component analysis model, or an entity embeddings model; and allowing, by the one or more processors, a user to select a product from the first UGC image.
With regard to claim 9, the prior art of record, alone or combined, neither anticipates nor renders obvious, a system reciting similar limitations.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ANNE MARIE GEORGALAS whose telephone number is (571)270-1258 E.S.T.. The examiner can normally be reached on Monday-Friday 8:30am-5:00pm.
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, Marissa Thein can be reached on 571-272-6764. 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.
/Anne M Georgalas/
Primary Examiner, Art Unit 3689
Prosecution Insights