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 Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 2, 7, 9, 11, 13-17 and 20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Bai et al. (U.S. Patent Application Publication No. 2025/0029289), referred herein as Bai.
Regarding claim 1, Bai teaches a computer-implemented method to generate multiple, self-consistent synthetic images of a visual object, the method comprising: obtaining, by a computing system (figs 1 and 2, computing system 100/200; paragraphs 21 and 32),
a textual prompt that textually describes the visual object; for each of one or more update iterations: processing, by the computing system, the textual prompt with a machine-learned image generation model to generate a plurality of synthetic images that depict the visual object; and training, by the computing system, the machine-learned image generation model on at least some of the plurality of synthetic images (paragraph 20, lines 1-10; paragraph 22, lines 8-13; paragraphs 24 and 34; paragraph 35, lines 1-5; a text prompt is obtained that describes an object, and for one or more update iterations, the text prompt is processed with an MLM to generate synthetic images and train the MLM using the images); and
after a final update iteration of the one or more update iterations, processing, by the computing system, the textual prompt with the machine-learned image generation model to generate a plurality of output images that depict the visual object (paragraph 20, the last 5 lines; paragraph 26; paragraph 40; paragraph 43, lines 1-12; once an objective is reached, the update iterations stop, and the text prompt is processed by the MLM to generate output images depicting the object).
Regarding claim 2, Bai teaches the computer-implemented method of claim 1, wherein training, by the computing system, the machine-learned image generation model on at least some of the plurality of synthetic images comprises: selecting, by the computing system, a subset of the plurality of synthetic images that exhibit visual cohesion; and training, by the computing system, the machine-learned image generation model on the selected a subset of the plurality of synthetic images that exhibit visual cohesion (paragraphs 55 and 57; paragraph 58, lines 1-14).
Regarding claim 7, Bai teaches the computer-implemented method of claim 1, wherein training, by the computing system, the machine-learned image generation model on at least some of the plurality of synthetic images comprises performing a text-to-image personalization technique on at least some of the plurality of synthetic images (paragraph 34; paragraph 35, lines 1-10; paragraph 40; paragraph 43, lines 1-12).
Regarding claim 9, Bai teaches the computer-implemented method of claim 1, wherein training, by the computing system, the machine-learned image generation model on at least some of the plurality of synthetic images comprises updating one or more parameters of the machine-learned image generation model (paragraph 20, lines 1-10; paragraph 22, lines 8-13; paragraph 24).
Regarding claim 11, Bai teaches the computer-implemented method of claim 1, wherein the one or more update iterations comprise a plurality of update iterations (paragraph 20, lines 1-10).
Regarding claim 13, Bai teaches the computer-implemented method of claim 1, wherein the visual object comprises a novel visual object not depicted in any training data on which the machine-learned image generation model has been trained (paragraph 37, lines 1-7).
Regarding claim 14, Bai teaches the computer-implemented method of claim 1, wherein the machine-learned image generation model comprises a pre-trained model (paragraphs 22 and 23; paragraph 33, the last 5 lines).
Regarding claim 15, Bai teaches the computer-implemented method of claim 1, wherein the machine-learned image generation model comprises a denoising diffusion model (paragraph 37, lines 1-11; paragraph 49, lines 1-8).
Regarding claim 16, the limitations of this claim substantially correspond to the limitations of claim 1 (except for the processors, media, and instructions, which are disclosed by Bai, fig 6; paragraph 112); thus they are rejected on similar grounds.
Regarding claim 17, the limitations of this claim substantially correspond to the limitations of claim 2; thus they are rejected on similar grounds.
Regarding claim 20, the limitations of this claim substantially correspond to the limitations of claim 1 (except for the media and system, which are disclosed by Bai, fig 6; paragraph 112); thus they are rejected on similar grounds.
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.
Claims 3-6, 18, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Bai, in view of Xie et al. (U.S. Patent Application Publication No. 2026/0073684), referred herein as Xie.
Regarding claim 3, Bai teaches the computer-implemented method of claim 2, wherein selecting, by the computing system, the subset of the plurality of synthetic images that exhibit visual cohesion comprises: generating, by the computing system, a plurality of embeddings respectively for the plurality of synthetic images in a latent embedding space, and selecting, by the computing system, one of the plurality of embeddings as the subset of the plurality of synthetic images based on the plurality of cohesion values (paragraph 35, the last 7 lines; paragraph 49).
Bai does not teach clustering, by the computing system, the plurality of embeddings into a plurality of clusters; evaluating, by the computing system, a cohesion measure for each of the plurality of clusters to determine a plurality of cohesion values respectively for the plurality of clusters; and selecting, by the computing system, one of the plurality of clusters based on the plurality of cohesion values.
However, in a similar field of endeavor, Xie teaches a computing system comprising a trained machine learning model for generating and analyzing image data by iteratively updating and clustering a plurality of feature embeddings into a plurality of clusters (paragraph 26, the last 6 lines; paragraph 39, lines 1-14; paragraphs 47 and 48), evaluating, by the computing system, a cohesion measure for each of the plurality of clusters to determine a plurality of cohesion values respectively for the plurality of clusters, and selecting, by the computing system, one of the plurality of clusters based on the plurality of cohesion values (paragraph 39, lines 1-14; paragraph 40, the last 11 lines; paragraph 41, lines 1-7; paragraph 46, lines 1-13; paragraph 47; paragraph 55).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the clustering of Xie with the text prompt MLM processing of Bai because this facilitates evaluation and updating based on clusters rather than individual descriptors, thereby improving both storage requirements and processing efficiency and accuracy (see, for example, Xie, paragraph 26).
Regarding claim 4, Bai in view of Xie teaches the computer-implemented method of claim 3, wherein the cohesion measure evaluated for each cluster comprises an average Euclidean distance between members of the cluster and a centroid of the cluster (Xie, paragraph 17, lines 1-9; paragraph 56, lines 1-20; the motivation to combine is similar to that discussed above in the rejection of claim 3).
Regarding claim 5, Bai in view of Xie teaches the computer-implemented method of claim 3, further comprising discarding, by the computing system, any cluster with a number of members below a threshold value (Xie, paragraphs 63 and 66; the motivation to combine is similar to that discussed above in the rejection of claim 3).
Regarding claim 6, Bai in view of Xie teaches the computer-implemented method of claim 3, wherein clustering, by the computing system, the plurality of embeddings into the plurality of clusters comprises performing a K-MEANS++ algorithm (Xie, paragraphs 51 and 69; the motivation to combine is similar to that discussed above in the rejection of claim 3).
Regarding claims 18 and 19, the limitations of these claims substantially correspond to the limitations of claims 3 and 4, respectively; thus they are rejected on similar grounds as their corresponding claims.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Bai, in view of Gal et al. (“An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion”; August 2022), referred herein as Gal.
Regarding claim 8, Bai teaches the computer-implemented method of claim 1, wherein training, by the computing system, the machine-learned image generation model on at least some of the plurality of synthetic images comprises performing reverse path learning (paragraph 37, lines 1-7).
Bai does not teach utilizing textual inversion to learn a set of dedicated learnable textual tokens.
However, in a similar field of endeavor, Gal teaches a system for obtaining and processing text prompts by a pre-trained machine learning model to generate synthetic images (page 2, the paragraph beginning “We propose…”; the paragraph beginning “To find…; page 4, the paragraph beginning “Our goal…”), and further comprising utilizing textual inversion to learn a set of dedicated learnable textual tokens (page 2, the paragraph beginning “To find…; page 4, fig 2 and its description; page 5, the section titled “Text embeddings”; the section titled “Textual inversion”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the textual inversion and tokens of Gal with the text prompt processing of Bai because this helps to more readily and accurately evaluate the text prompts such that higher quality, more accurate, and/or novel objects and images can be generated (see, for example, Gal, page 2, the paragraph beginning “We demonstrate…”; page 8, fig 5 and its description; pages 8-9, the paragraph beginning “The images…”).
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Bai, in view of Neerukonda et al. (U.S. Patent Application Publication No. 2025/0156644), referred herein as Neerukonda.
Regarding claim 10, Bai teaches the computer-implemented method of claim 9, wherein updating the one or more parameters of the machine-learned image generation model comprises learning and updating parameter values (paragraph 20, lines 1-10).
Bai does not teach learning a set of low-rank adaptation values.
However, in a similar field of endeavor, Neerukonda teaches a system for obtaining and processing text prompts by a pre-trained machine learning model to generate synthetic output (figs 1 and 2; paragraph 28; paragraph 31, lines 1-16; paragraph 38), comprising learning a set of low-rank adaptation values (paragraph 13, lines 1-13; paragraph 29, the last 12 lines).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine to low-rank adaptation updating of Neerukonda with the parameter updating of Bai because this can improve the tuning of the models and thus improve the quality of the synthetic training data training the machine learning model, and by extension, its output (see, for example, Neerukonda, paragraph 13, the last 18 lines; paragraph 20).
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Bai, in view of Liu et al. (U.S. Patent Application Publication No. 2016/0189003), referred herein as Liu.
Regarding claim 12, Bai teaches the computer-implemented method of claim 11, wherein the method comprises performing update iterations until a model convergence metric is satisfied (paragraph 20, lines 3-10; paragraph 25).
Bai does not teach that the convergence metric comprises an average pairwise Euclidean distance between the synthetic images being smaller than a predefined threshold.
However, in a similar field of endeavor, Liu teaches a computing system comprising training a machine learning model for analyzing cohesion of features input images to determine a convergence metric (figs 2; paragraphs 37 and 100), wherein the convergence metric comprises an average pairwise Euclidean distance between the synthetic images being smaller than a predefined threshold (paragraphs 95 and 96; paragraph 101).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the convergence metric of Liu with that of Bai because this facilitates better identification of the convergence, which thereby produces better results and improves the user experience (see, for example, Liu, paragraph 58, the last 9 lines; paragraph 100).
Conclusion
The following prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
O'Brien (U.S. Patent Application Publication No. 2021/0406698); Identifying anomalous activity from thermal images.
Admasu (U.S. Patent Application Publication No. 2023/0153988); System and method for identifying and analyzing desired properties of a material.
Harikumar (U.S. Patent No. 12,493,937); Prior guided latent diffusion.
Willmott (U.S. Patent Application Publication No. 2025/0005918); System and method for prompt searching.
Karpman (U.S. Patent No. 11,995,803); Training and deployment of image generation models.
Hinz (U.S. Patent Application Publication No. 2024/0320872); Image generation using a text and image conditioned machine learning model.
Hinz (U.S. Patent Application Publication No. 2024/0320873); Text-based image generation using an image-trained text.
Montalt (U.S. Patent Application Publication No. 2024/0378503); Generating synthetic training data.
Chen (U.S. Patent Application Publication No. 2025/0117973); Style-based image generation.
Handa (U.S. Patent Application Publication No. 2026/0154857); Style-aligned object image generation.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DAVID T WELCH whose telephone number is (571)270-5364. The examiner can normally be reached Monday-Thursday, 8:30-5:30 EST, and alternate Fridays, 9:00-2:30 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, Xiao Wu can be reached at 571-272-7761. 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.
DAVID T. WELCH
Primary Examiner
Art Unit 2613
/DAVID T WELCH/Primary Examiner, Art Unit 2613
Prosecution Insights