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 .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 7/7/2026 has been entered.
Response to Amendment
Claims 1-5, 7-20 are pending.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1-5, 7-20 have been considered but are moot because the new ground of rejection is directed towards the newly amended claim limitations that change the scope of the claims as a whole and are open to new grounds of rejection/interpretation.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-4, 7-8, 10-13, 15-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Willis et al. (US 20230104396) in view of Li et al. (US 20250285343) and Wen et al. “Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning and Discovery.”
Re claim 1, Willis teaches a method, comprising:
receiving, by a processing device, a request initiated by a user to generate a thumbnail image to be associated with a collection of media items stored by a content platform ([0063] FIG. 6 is a schematic flow chart diagram of a method 600 for generating a unique code for accessing a data collection. The method 600 includes authenticating at 602 a user and enabling the user to log in to an administrator portal of an account supported by the media server. The method 600 includes receiving data at 604 uploaded by the user and storing the data on the image server for cloud-based access. The method 600 includes providing at 606 a user interface to the user that enables the user to organize and manage the data collection. The method 600 includes receiving at 608 a thumbnail selection approval from the user, wherein the thumbnail comprises a screen capture from one or more data objects within the data collection. The method 600 includes generating at 610 a unique code that provides instructions to a personal device to access the data collection. The method 600 includes merging at 612 the unique code and the thumbnail selection to generate a scannable thumbnail.
identifying one or more stylistic features specified by the user ([0064] The data uploaded by the user may include one or more independent data objects such as videos, images, written works, numerical data, historical data, and so forth. The one or more data objects may be grouped together to generate a data collection. The data collection may be grouped together by the neural network 114 as described herein and/or manually grouped together by the user. One data collection may include one or more data objects that are related based on geographical location, time, quality, subject matter, or some other metric. The data collection may include one or more data objects that are stitched together on the media library 112 using common metadata to indicate that each of the one or more data objects should be associated with the same data collection).
generating, using the one or more stylistic features, a textual prompt that describes the thumbnail image to be generated (see [0074-0076], in reference to Fig. 8E-IG, wherein submit tags to be associated with the video 806 and these tags represent metadata for the video 806. The tags may, for example, who is depicted in the video, the subject of the video, the time or season the video was captured, and so forth. The tags will be stored in association with the video 806 on the media library 112….FIG. 8F illustrates wherein the unique code 810 is displayed in the center of the thumbnail 808. The combination of the unique code 810 and the thumbnail 808 produces a scannable thumbnail. The scannable thumbnail may be captured or scanned with a sensor and provide instructions to redirect to a website, file system, database, and so forth. The unique code 810 may provide instructions to access a website where the video 806 may be viewed…The thumbnail 808 may include an image frame from a video 806, a screen capture, an image, a document, a graphic, and so forth. The unique code 810 may redirect a computing device to access additional media associated with the thumbnail 808).
Willis further teaches obtaining one or more outputs, the one or more outputs specifying respective one or more thumbnail images (see [0045], generate a scannable thumbnail by merging unique code with a selected thumbnail).
Willis does not explicitly teach generating, using the one or more stylistic features, a textual prompt for an AI generative model, wherein the textual prompt generated for the AI generated model using the one or more stylistic features describes the thumbnail image to be generated, causing an AI generative model to process the textual prompt describing the thumbnail to be generated; and obtaining one or more outputs from the AI generative model, the one or more outputs specifying respective one or more images.
However, LI teaches generating, using the one or more stylistic features, a textual prompt for an AI generative model wherein the textual prompt generated for the AI generated model using the one or more stylistic features describes the thumbnail image to be generated
([0054] FIG. 1D is an overall system diagram, according to one embodiment. The system provides users with an easy starting point, allowing them to easily create an avatar using subject image(s) 150, a style request 152, and optionally an object request 153. The style request 152 can be at least one of a user-entered style text 152a, a user-selected style image 152b1 (e.g., from a style image library), or a user-uploaded style image 152b2), [0055] The system then uses the LMM 126a to convert any non-textual input(s) into textural description(s), and then rewrites all textural descriptions into an image generation prompt for the LVM 126b to create personalized avatar(s). FIG. 1D shows the avatar with Christmas theme flat design style & objects 154d, and a variation. For instance, the LMM 126a interprets the subject image(s) 150 and extracts key identifying features of the subjects in step 1. In step 2, the LMM 126a interprets the user-uploaded style image 152b2 and extracts key style feature description, including lighting, angle, theme, colors, etc. Alternatively or concurrently, the LMM 126a (1) directly incorporate or interpret the user-entered style text 152a, and/or (2) loads a pre-defined textual style description based on the user-selected style image 152b1 from the library. In step 3, the LMM 126a combines all of the textual descriptions from steps 1 and 2, and rewrites the combined textual descriptions into an image generation prompt).
causing an AI generative model to process the textual prompt describing the thumbnail to be generated; and obtaining one or more outputs from the AI generative model, the one or more outputs specifying respective one or more images (see [0055], in reference to Fig. 1D, wherein the LMM combines text outputs from steps 1 and 2, and rewrites into an image prompt, and one or more AI generated images as outputs.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Willis’s method of generating a thumbnail image to explicitly include AI generative model to process user specified stylistic features into textual prompts to describe an image to be generated, and output one or more outputs specifying respective images, as taught by Li, as the references are in the analogous art of image generation from user specified stylistic features. An advantage of the modification is that it achieves the result of using AI models to process textual prompts for one or more outputs specifying images based on identified features specified by a user including preferences and styles.
Willis and Li do not explicitly teach determining based on an allowed list of prompt terms, whether the textual prompt satisfies a content appropriateness condition, and responsive to determining that the textual prompt satisfies the content appropriateness condition, causing the AI generative model to process the textual prompt describing the thumbnail image to be generated.
However, Wen teaches determining based on an allowed list of prompt terms, whether the textual prompt satisfies a content appropriateness condition, and responsive to determining that the textual prompt satisfies the content appropriateness condition, causing the AI generative model to process the textual prompt describing the thumbnail image to be generated (see p. 8-9, 6. Safety Concerns, wherein token or word-level filters are often used in text-to-image diffusion model APIs to prevent the generation of NSFW or copyrighted content. For instance, image generation API Midjourney has banned prompts containing the substring “Afghan” due to copyright issue with the famous photo of an Afghan girl,) and (see p. 9: However, prompt optimization can be used as a mechanism to bypass simple rule-based content filters. PEZ can generate a prompt that avoids banned tokens, yet still matches
textual features with the original target prompt “Afghan girl.” Figure 7 shows the output of Midjourney using an optimized prompt which successfully reproduces the banned
image without containing the banned word “Afghan.” Note that the prompt seems to incorrectly associate the subject of the image, Sharbat Gula, with the Taliban). Wen teaches determining based on an allowed list of prompt terms, whether the textual prompt satisfies a content appropriateness condition (terms that do not contain “Afghan girl” are appropriate)
and responsive to determining that the textual prompt satisfies the content appropriateness condition, causing the AI generative model to process the textual prompt describing the thumbnail image to be generated (generating image without banned word).
Willis, Li, and Wen teaches claim 1. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Willis and Li’s method of generating a thumbnail image using textual prompts, to explicitly include determining whether the textual prompt satisfies a content appropriateness condition based on an allowed list and processing the textual prompt based on satisfying the condition, as taught by Wen, as the references are in the analogous art of image generation from user textual prompts. An advantage of the modification is that it achieves the result of using AI models to process textual prompts based on allowable list of prompt terms, to minimize unsafe/offensive generated content.
Re claim 2, Willis and Li and Wen teaches claim 1. Furthermore, Li teaches causing a list of stylistic features to be presented via a UI and receiving, via the UI, one or more selections of respective one or more stylistic features from the list of stylistic features (see [0070-0071], in reference to Fig. 2c, wherein a user selects one of the following images as a style image for generating avatar). For motivation, see claim 1.
Re claim 3, Willis and Li and Wen teaches claim 2. Furthermore, Li teaches wherein the list of stylistic features is personalized based on a user profile associated with a user [[0021] Another technical benefit of the approach provided herein is to capitalize on a suite of powerful tools to understand user subject requests and style requests, both of which can be in image and/or text format. By converting a style image into a text style prompt, the system generates the avatar for the subject in the subject request in the user-desired style, and optionally creates a background with object(s) based on an object request. Therefore, the generated avatar(s) more accurately represents the user preferences. Not only does this improve the productivity of the user, but this approach can also decrease the computing resources required to refine the style based on refined user queries to the generative models), ([0030] The term “avatar” refers to a visual representation of a person or character for use in digital contexts. It's usually a computer-generated image, such as a bitmoji. On social media, the term “avatar” also refers to a profile image represents a user on the platform), ([0061] In some implementations, the system makes the avatar 154 produced by the pipeline editable, such as adding textual content in the avatar 154, thus offering more user control over their AI-generated content (AIGC) experiences. For instance, after generating the avatar 154, either the LMM 126a or the prompt construction unit 124 can query the user for more usage context details, such as the nature of the platform/application to use the avatar, and then add more details to the avatar 154. For example, if the user applies the avatar in a sports fan chat room, the system can add more visual elements associated with the supported sports teams to the user avatar. In another embodiment, the system can extract/infer the avatar usage context details from the user database 128. For example, the prompt construction unit 124 can retrieve user preference data 128a (shown in FIG. 3) from the user database 128 based on an indication identifying the user. The indication may be a user identifier (e.g., a username, email address, and the like), and/or other identifier associated with the user that the application services platform 110 can use to identify the user and retrieve user data. The user data can include a username, a user organization, a user preferred graphic design style (e.g., minimalism, retro, art deco, Memphis design, Swiss style, Bauhaus, pop art, punk, etc.), and the like. As such, when the user does not provide the avatar usage context details, the prompt construction unit 124 may retrieve the missing information from the user preference data 128a, instead of asking more questions for the missing information via an AI chat interface), ([0069] The Generate button 215b can be selected to generate an avatar with desired style(s) and an optional background with object(s) corresponding to a user style request (e.g., the style request 152) and a user object request (e.g., a birthday cake and balloons). The Share button 215c can be selected to trigger a dropdown list of applications to share an avatar (e.g., the avatar 154). For example, the user can post the avatar on a social media application (e.g., Facebook®) to celebrate the user's birthday. The search field 215d is for a user to enter a search word, phrase, paragraph, and the like within the visual content library 142, the requests, prompts, and responses 144, the extracted/inferred user data 146 (e.g., activities, preferences, or the like), the other asset data 148, and the like. The fields in the visual style transfer to avatar application can provide auto-fill and/or spell-check functions), [0095] The above-discussed visual content library 142 (storing e.g., subjects, styles, objects, backgrounds, or the like), request, prompts and responses 144, extracted/inferred user data 146 (e.g., user preferences), and other asset data 148 can be stored in the enterprise data storage 140. The extracted/inferred user data 146 (e.g., activities, preferences, or the like) is tentatively linked with a user ID during a user session and saved in a cache. After the user session, extracted/inferred user data 146 is de-linked form the user ID as metadata of the resulted new style image(s) and saved in the visual content library 142. In addition, the extracted/inferred user data 146 linked with the user ID is saved back to the user database 128). For motivation, see claim 1.
Re claim 4, Willis and Li and Wen teaches claim 2. Furthermore, Li teaches wherein the list of stylistic features is translated to a natural language specified by a user profile associated with the user ([0048] The system provides users with the ability to generate personalized avatars with an unlimited range of styles. This is made possible through: natural language style descriptions and/or style image upload. With the natural language style descriptions, users can simply describe the desired style in their own words, and the system will generate a personalized avatar for the user with a style that matches the description. Alternatively, the user can also upload an image as a style goal. This feature enables the user to take inspiration from any images and use them as a basis for generating new and unique avatars). For motivation, see claim 2.
Re claim 7, Willis, Li, and Wen teaches claim 1. Furthermore, Wen teaches wherein determining whether the textual prompt satisfies the content appropriateness condition further comprises determining whether one or more terms of the textual prompt are included on the allowed list of prompt terms (see p. 8-9, 6. Safety Concerns, wherein token or word-level filters are often used in text-to-image diffusion model APIs to prevent the generation of NSFW or copyrighted content. For instance, image generation API Midjourney has banned prompts containing the substring “Afghan” due to copyright issue with the famous photo of an Afghan girl,) and (see p. 9: However, prompt optimization can be used as a mechanism to bypass simple rule-based content filters. PEZ can generate a prompt that avoids banned tokens, yet still matches textual features with the original target prompt “Afghan girl.” Figure 7 shows the output of Midjourney using an optimized prompt which successfully reproduces the banned image without containing the banned word “Afghan.” Note that the prompt seems to incorrectly associate the subject of the image, Sharbat Gula, with the Taliban). For motivation, see claim 1. For motivation, see claim 1.
Re claim 8, Willis and Li and Wen teach claim 1. Furthermore, Willis teaches receiving, via a UI, an input identifying a chosen thumbnail image of the one or more thumbnail images; and associating the chosen thumbnail image with the collection of media items ([0063] FIG. 6 is a schematic flow chart diagram of a method 600 for generating a unique code for accessing a data collection. The method 600 includes authenticating at 602 a user and enabling the user to log in to an administrator portal of an account supported by the media server. The method 600 includes receiving data at 604 uploaded by the user and storing the data on the image server for cloud-based access. The method 600 includes providing at 606 a user interface to the user that enables the user to organize and manage the data collection. The method 600 includes receiving at 608 a thumbnail selection approval from the user, wherein the thumbnail comprises a screen capture from one or more data objects within the data collection. The method 600 includes generating at 610 a unique code that provides instructions to a personal device to access the data collection. The method 600 includes merging at 612 the unique code and the thumbnail selection to generate a scannable thumbnail) and ([0064] The data uploaded by the user may include one or more independent data objects such as videos, images, written works, numerical data, historical data, and so forth. The one or more data objects may be grouped together to generate a data collection. The data collection may be grouped together by the neural network 114 as described herein and/or manually grouped together by the user. One data collection may include one or more data objects that are related based on geographical location, time, quality, subject matter, or some other metric. The data collection may include one or more data objects that are stitched together on the media library 112 using common metadata to indicate that each of the one or more data objects should be associated with the same data collection). For motivation, see claim 1.
Claim 10 claims limitations in scope to claim 1, and is rejected for at least the reasons above.
Claims 11-13 claim limitations in scope to claims 2-4 and is rejected for at least the reasons above.
Claim 15 claims limitations in scope to claim 7, and is rejected for at least the reasons above.
Claim 16 claims limitations in scope to claim 1, and is rejected for at least the reasons above.
Claims 17-19 claim limitations in scope to claims 2-4 and is rejected for at least the reasons above.
Claim(s) 5, 14, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Willis et al. (US 20230104396) in view of Li et al. (US 20250285343), Wen et al. “Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning and Discovery,” and Voss et al. (US 11776194).
Re claim 5, Willis, and Li and Wen teaches claim 2.
Willis, and Li and Wen do not explicitly teach receiving, via the UI, a list refresh command, and response to receiving the list refresh command, refreshing the list of stylistic features.
However, Voss teaches receiving, via the UI, a list refresh command, and response to receiving the list refresh command, refreshing the list of stylistic features (abstract: Aspects of the present disclosure involve a system comprising a computer-readable storage medium storing a program and method for animating a pull-to-refresh gesture. The program and method provide for receiving a pull gesture in a messaging application; selecting, in response to receiving the pull gesture, a set of users corresponding to contacts in the messaging application; and displaying a set of images for each user in the set of users, in association with refreshing screen content) and (see col 2, line 10-24: In some cases, a user may wish to refresh a screen being displayed by the messaging application. For example, the user may perform a pull-to-refresh gesture on a list of users presented on the device screen. The pull-to-refresh gesture includes the user touching the screen, pulling the screen downward past a threshold pull distance, and then releasing the touch. In response to detecting the pull portion of the gesture, a messaging application typically provides an animation depicting the top edge of the screen being pulled down along with the pull gesture. In response to the release portion, the messaging application provides an animation for the screen to move back upwards, and for display of the updated content (e.g., an updated list of users). In some cases, an end user may wish to have a more engaging experience with respect to refreshing screen content).
Willis, Li, Wen, and Voss teach claim 5. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Willis, Li, and Wen’s method of displaying a list of screen content, such as styles, to explicitly refresh the list, as taught by Li, as the references are in the analogous art of display of a list of content data to a user. An advantage of the modification is that it achieves the result of using a user interface of with a refresh command to update the list displayed to a user to provide for an updated list of styles to choose from.
Claims 14 claims limitations in scope to claim 5 and is rejected for at least the reasons above.
Claims 20 claims limitations in scope to claim 5 and is rejected for at least the reasons above.
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Willis et al. (US 20230104396) in view of Li et al. (US 20250285343), Wen et al. “Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning and Discovery, ” and Knipfing et al. (US 20240354555).
Re claim 9, Willis and Li, and Wen teaches claim 1. Willis and Li and Wen do not explicitly teach providing each of the one or more thumbnail images as input to a second trained AI model; and obtaining one or more outputs of the second trained AI model, the one or more outputs of the second AI model indicating a probability of the thumbnail image comprising an inappropriate content.
However, Knipfing teaches providing each of the one or more thumbnail images as input to a second trained AI model; and obtaining one or more outputs of the second trained AI model, the one or more outputs of the second AI model indicating a probability of the thumbnail image comprising an inappropriate content (0056] A collection management system 220 is operationally responsible for managing sets or collections of media (e.g., collections of text, image video, and audio data). A collection of content (e.g., messages, including images, video, text, and audio) may be organized into an “event gallery” or an “event story.” Such a collection may be made available for a specified time period, such as the duration of an event to which the content relates. For example, content relating to a music concert may be made available as a “story” for the duration of that music concert. The collection management system 220 may also be responsible for publishing an icon that provides notification of a particular collection to the user interface of the interaction client 104. The collection management system 220 includes a curation function that allows a collection manager to manage and curate a particular collection of content. For example, the curation interface enables an event organizer to curate a collection of content relating to a specific event (e.g., to delete inappropriate content or redundant messages). Additionally, the collection management system 220 employs machine vision (or image recognition technology) and content rules to curate a content collection automatically. In certain examples, compensation may be paid to a user to include user-generated content into a collection. In such cases, the collection management system 220 operates to automatically make payments to such users to use their content). To clarify, Knipfing teaches indicating a probability that content is inappropriate or redundant (such as 100% or 0%) and deleting the indicated inappropriate content to curate the content collection automatically using machine vision.
Willis, Li, Wen, and Knipfing teaches claim 9. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Willis, Wen, and Li’s method of processing image data of thumbnail images, to explicitly include indicating that image content of an image as inappropriate content, as taught by Knipfing, as the references are in the analogous art of image processing in an AI model. An advantage of the modification is that it achieves the result of filtering and deleting image content deemed as inappropriate and keeping content that is appropriate.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Peter Hoang whose telephone number is (571)270-1346. The examiner can normally be reached Monday-Friday 8:00 am - 5:00 pm PST.
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, Hajnik F. Daniel can be reached at (571) 272-7642. 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.
/PETER HOANG/ Primary Examiner, Art Unit 2616
Prosecution Insights