Prosecution Insights
Last updated: November 20, 2026
Application No. 19/036,614

GENERATIVE MODEL AND LATENT FRAME APPROXIMATION FOR MEDIA DATA GENERATION

Non-Final OA §103
Filed
Jan 24, 2025
Examiner
RICHER, AARON M
Art Unit
2617
Tech Center
2600 — Communications
Assignee
Qualcomm Incorporated
OA Round
1 (Non-Final)
52%
Grant Probability
Moderate
1-2
OA Rounds
1y 11m
Est. Remaining
73%
With Interview

Examiner Intelligence

Grants 52% of resolved cases
52%
Career Allowance Rate
253 granted / 485 resolved
-9.8% vs TC avg
Strong +20% interview lift
Without
With
+20.4%
Interview Lift
resolved cases with interview
Typical timeline
3y 9m
Avg Prosecution
24 currently pending
Career history
517
Total Applications
across all art units

Statute-Specific Performance

§101
10.0%
-30.0% vs TC avg
§103
54.4%
+14.4% vs TC avg
§102
12.8%
-27.2% vs TC avg
§112
20.1%
-19.9% vs TC avg
Black line = Tech Center average estimate • Based on career data from 485 resolved cases

Office Action

§103
DETAILED ACTION Notice of Pre-AIA or AIA Status The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . Claim 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. This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention. Claims 1, 2, 6, 8, 9, 12, 13, 15-17, 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Wang (U.S. Publication 2023/0147641) in view of Mosseri (U.S. Publication 2025/0238905). As to claim 1, Wang discloses a device comprising a memory configured to store a generative model and one or more processors (p. 2, section 0047; p. 42, section 0355; p. 43-44, section 0364; p. 48, sections 0394-0395) configured to: obtain an input image frame (p. 2-3, section 0049; p. 3, section 0054; an object from a single image frame is chosen to be animated); generate, based on the input image frame, a time sequence of multiple latent image frames (p. 2, sections 0047-p. 3, section 0051; a time sequence of frames is generated in latent space); for a first sampling operation of multiple sampling operations: receive a first input version of the multiple latent image frames; perform an operation, based on the generative model, on a first subset of the first input version of the multiple latent image frames to generate a first output subset of the multiple latent image frames (p. 2, section 0047; p. 3, sections 0053-0054; multiple sampling operations are performed on the input latent image frames to produce subsets of output frames); output a first output version of the multiple latent image frames, the first output version including the first output subset and a second output subset; and output, based on the multiple sampling operations, multiple output image frames associated with the input image frames (p. 3, sections 0051-0054; based on various augmentation sampling operations on input frames, multiple output frames including different subsets associated with different characteristics are output). Wang does not disclose, but Mosseri discloses performing a diffusion operation, based on the generative model, on a first subset of the first input version of the multiple latent image frames to generate a first output subset of the multiple latent image frames, and generating a second output subset of the multiple latent image frames based on the first input version of the multiple latent image frames and based on the first output subset of the multiple latent image frames (fig. 2; fig. 4, p. 4, sections 0044-0046; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; a diffusion model/operation is used to generate output from the input latent frame representations; as can be seen in the figures, a second output subset such as element 212 is generated from a first input 204 and a first output subset such as 210 that is generated based on a first input 204; similar relationships can be found elsewhere in the figures). The motivation for this is to optimize memory usage while preserving motion consistency of generated video (p. 1, section 0004). It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to modify Wang to perform a diffusion operation, based on the generative model, on a first subset of the first input version of the multiple latent image frames to generate a first output subset of the multiple latent image frames, and generate a second output subset of the multiple latent image frames based on the first input version of the multiple latent image frames and based on the first output subset of the multiple latent image frames in order to optimize memory usage while preserving motion consistency of generated video as taught by Mosseri. As to claim 2, Wang discloses wherein the generative model includes an image-to-video generative model (p. 2-3, section 0049; a single image is used to generate the animated video) and the multiple output image frames is a time sequence of the multiple output image frames (p. 2, sections 0047-p. 3, section 0051; a time sequence of frames is generated with particular time stamps). As to claim 6, Wang does not disclose, but Mosseri discloses wherein the one or more processors are configured to, for a second sampling operation of the multiple sampling operations: receive a second input version of the multiple latent image frames, perform the diffusion operation, based on the generative model, on the second input version of the multiple latent image frames to generate a second output version of the multiple latent image frames; and output the second output version of the multiple latent image frames (fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; a diffusion model/operation is used to generate output from the input latent frame representations; the process is iterative, meaning that before or after a first input version is operated on to generate a first output version, a second input version is operated on to generate a second output version, and so on for a particular number of time steps). Motivation for the combination is given in the rejection to claim 1. As to claim 8, Wang does not disclose, but Mosseri discloses wherein the second sampling operation is performed prior to the first sampling operation; and the second output version is provided as the first input version for the second sampling operation (fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; a diffusion model/operation is used to generate output from the input latent frame representations; the process is iterative, meaning that before or after a first input version is operated on to generate a first output version, a second input version is operated on to generate a second output version, and so on for a particular number of time steps; each iteration would have input equal to the previous output, so the second output version would be the first input version if the second operation were the iteration previous to the first). Motivation for the combination is given in the rejection to claim 1. As to claim 9, Wang does not disclose, but Mosseri discloses wherein the second sampling operation is performed after the first sampling operation; and the first output version is provided as the second input version for the second sampling operation (fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; a diffusion model/operation is used to generate output from the input latent frame representations; the process is iterative, meaning that before or after a first input version is operated on to generate a first output version, a second input version is operated on to generate a second output version, and so on for a particular number of time steps; each iteration would have input equal to the previous output, so the first output version would be the second input version if the first operation were the iteration previous to the second). As to claim 12, Wang discloses wherein the one or more processors are configured to encode the input image frame to generate a latent representation of the input image frame (p. 2, section 0049-p. 3, section 0050; a single image is the input and each image is represented by a latent representation); and decode a final version of the multiple latent image frames based on the multiple sampling operations to generate the multiple output image frames and wherein the multiple output image frames include image frames associated with the input image frame (p. 3, sections 0052-0054; the multiple output frames produced from the single image are used to finally synthesize a video, necessitating decoding from latent space). Wang does not disclose, but Mosseri discloses encoding, via a variational autoencoder (VAE) (p. 9, section 0116) and that the frames include fourteen or more image frames (p. 2, section 0026). Motivation for the combination is given in the rejection to claim 1. As to claim 13, Wang discloses wherein the generative model is applied to perform a text-based video generation operation, a text-based video content editing operation, image-based video generation operation, a video enhancement operation, video compression, a data augmentation operation, or a combination thereof (p. 2, section 0049-p. 3, section 0053; a video is produced from an image reading on an image-based video generation operation; further, the video is enhanced and augmented). As to claim 15, Wang discloses a device further comprising a display device coupled to the one or more processors and configured to output the multiple output image frames as video content (fig. 5, element 506; fig. 9, element 924; fig. 10, element 1006; fig. 13, element 1345; fig. 16, elements 1610a/b; fig, 23, element 2311; p. 1, section 0045-p. 2, section 0046; the video sequence is output to a display device). As to claim 16, Wang discloses a device further comprising a modem coupled to the one or more processors, the modem configured to transmit the multiple output image frames to a second device for output by the second device (fig. 16; p. 16, section 0172-p. 17, section 0173; frames are output to a remote second display device; the communication link between devices would involve modulation/demodulation of a signal and would read on a modem) As to claim 17, Wang discloses a device further comprising a microphone configured to provide an input signal to the one or more processors (p. 9, section 0094; p. 9, section 0099). As to claim 19, see the rejection to claim 1. As to claim 20, see the rejection to claim 1. Further, Wang discloses a non-transitory computer-readable medium that stores instructions that are executable by one or more processors to cause the one or more processors to perform the method (p. 47, section 0389). Claims 3-5, 10 and 11 are rejected under 35 U.S.C. 103 as being unpatentable over Wang in view of Mosseri and further in view of Besenbruch (U.S. Publication 2022/0279183). As to claim 3, Wang does not disclose, but Besenbruch discloses wherein, to generate the second output subset of the multiple latent image frames, the one or more processors are configured to interpolate the second output subset based on the first input version and the first output subset (p. 4, sections 0103-0113; using latent representations, a weighted sum/interpolation is done between output and input versions to backpropagate a loss and create a second output frame). The motivation for this is to obtain trained neural networks that can provide good quality output images (p. 25, section 0831). It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to modify Wang and Mosseri to interpolate the second output subset based on the first input version and the first output subset in order to obtain trained neural networks that can provide good quality output images as taught by Besenbruch. As to claim 4, Wang does not disclose, but Besenbruch discloses wherein, to interpolate the second output subset, the one or more processors are configured to: determine a change value based on the first subset and the first output subset (p. 4, sections 0103-0113; the loss function is a measure of change between the output and training image); determine a weight value (p. 4, sections 0103-0113; p. 25, section 0839; a weight value for the loss is assigned, necessitating a determination of the value); modify the change value based on the weight value (p. 4, sections 0103-0113; p. 25, section 0839; the change/loss value is multiplied by the weight, modifying the change/loss value); and generate the second output subset based on the modified change value and a second subset of the first input version of the multiple latent image frames (p. 4, sections 0103-0113; a weighted sum/interpolation is done between output and input versions to backpropagate a loss and create the second output frame). Motivation for the combination is given in the rejection to claim 3. As to claim 5, Besenbruch discloses wherein, to interpolate the second output subset, the one or more processors are configured to perform a linear interpolation operation (p. 4, sections 0103-0113; a weighted sum, which is a linear interpolation operation, is calculated). Motivation for the combination is given in the rejection to claim 3. As to claim 10, Wang discloses outputting a third output version of the multiple latent image frames, the third output version including the third output subset and the fourth output subset (p. 3, sections 0051-0054; based on various augmentation sampling operations on input frames, multiple output frames including different subsets associated with different characteristics are output). Wang does not disclose, but Mosseri discloses for a third sampling operation of the multiple sampling operations: receive the second output version of the multiple latent image frames as a third input version of the multiple latent image frames; perform the diffusion operation, based on the generative model, on a first subset of the third input version of the multiple latent image frames to generate a third output subset of the multiple latent image frames (fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; a diffusion model/operation is used to generate output from the input latent frame representations; the process is iterative meaning that before or after a third input version is operated on to generate a third output version, a second input version is operated on to generate a second output version, and so on for a particular number of time steps; each iteration would have input equal to the previous output, so the second output version would be the third input version if the second operation were the iteration previous to the third). Motivation for the combination is given in the rejection to claim 1. Wang in view of Mosseri does not disclose, but Besenbruch discloses generating a fourth output subset of the multiple latent image frames based on the third input version of the multiple latent image frames and based on the third output subset of the multiple latent image frames; and output a third output version of the multiple latent image frames (p. 4, sections 0103-0113; using latent representations, a weighted sum/interpolation is done between output and input versions to backpropagate a loss and create a second output frame; these would read on the third output, third input, and fourth output). Motivation for the combination of references is given in the rejection to claim 3. As to claim 11, Wang does not disclose, but Mosseri discloses wherein the multiple latent image frames include a subset of latent image frames; and each of the first subset of the first input version and the first subset of the third input version are associated with the subset of latent image frames of the multiple latent image frames (fig. 2; fig. 4, p. 4, sections 0044-0046; p. 4-5, section 0056; p. 5, sections 0059-0068; p. 5-6, section 0074; p. 6, section 0081; p. 7, section 0094-p. 8, section 0103; a diffusion model/operation is used to generate output from the input latent frame representations; the process is iterative such that each subsequent input is the previous iteration’s output and each set of frames is an iterated version of the original frames, thus associating a second output/third input with a first input and an original set of frames). Motivation for the combination of references is given in the rejection to claim 1. Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Wang in view of Mosseri and further in view of Varnica (U.S. Patent 8,533,496). As to claim 7, Wang does not disclose, but Varnica discloses a first power consumption associated with performance of the first sampling operation is less than a second power consumption associated with performance of the second sampling operation (col. 6, lines 19-29; a first sampling operation converging quickly has less power consumption than a second converging less quickly). The motivation for this is to account for different signal-to-noise ratios. It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to modify Wang and Mosseri to have a first power consumption associated with performance of the first sampling operation be less than a second power consumption associated with performance of the second sampling operation in order to account for different signal-to-noise ratios as taught by Varnica. Claim 14 is rejected under 35 U.S.C. 103 as being unpatentable over Wang in view of Mosseri and further in view of Luengo Muntion (U.S. Publication 2026/0301380). As to claim 14, Wang does not disclose, but Luengo Muntion discloses one or more cameras coupled to the one or more processors and configured to generate image data associated with the input image frame; and an input device configured to receive an input and provide the input to the one or more processors, wherein the input includes a request to generate video data including the multiple output image frames based on the image data from the one or more cameras (p. 4, section 0043-p. 5, section 0046; p. 6, section 0056; synthetic frame data is generated based on an input user request associated with an input camera image frame). The motivation for this is to better train models for downstream tasks (p. 2, sections 0025-0027). It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to modify Wang and Mosseri to generate image data associated with the input image frame and receive an input and provide the input to the one or more processors, wherein the input includes a request to generate video data including the multiple output image frames based on the image data from the one or more cameras in order to better train models for downstream tasks as taught by Luengo Muntion. Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Wang in view of Mosseri and further in view of Caba Heilbron (U.S. Publication 2026/0094440). As to claim 18, Wang does not disclose, but Caba Heilbron discloses the one or more processors are configured to perform a speech to text conversion operation on the input to generate text data (p. 2, section 0016); and the generative model is applied, based on the text data, to perform a text-based video generation operation or a text-based video content editing operation (p. 1, section 0003; p. 1, section 0015-p. 2, section 0016; natural language text is used to assemble/generate or edit a video). The motivation for this is to increase accessibility. It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to modify Wang and Mosseri to perform a speech to text conversion operation on the input to generate text data and perform a text-based video generation operation or a text-based video content editing operation in order to increase accessibility as taught by Caba Heilbron. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to AARON M RICHER whose telephone number is (571)272-7790. The examiner can normally be reached 9AM-5PM. 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, King Poon can be reached at (571)272-7440. 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. /AARON M RICHER/Primary Examiner, Art Unit 2617
Read full office action

Prosecution Timeline

Jan 24, 2025
Application Filed
Oct 15, 2026
Non-Final Rejection mailed — §103
Nov 17, 2026
Interview Requested

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12762775
REAL-TIME ANIMATION MOTION CAPTURE
5y 4m to grant Granted Oct 13, 2026
Patent 12755492
AUGMENTED REALITY OR VIRTUAL REALITY SYSTEM WITH ACTIVE LOCALISATION OF TOOLS, USE AND ASSOCIATED PROCEDURE,
3y 8m to grant Granted Oct 06, 2026
Patent 12755394
MODELING SECONDARY MOTION BASED ON THREE-DIMENSIONAL MODELS
3y 10m to grant Granted Oct 06, 2026
Patent 12749146
GENERATING IMAGE BLENDING WEIGHTS
4y 9m to grant Granted Sep 29, 2026
Patent 12743852
Prediction of Mechanical Properties of Sedimentary Rocks based on a Grain to Grain Parametric Cohesive Contact Model
2y 5m to grant Granted Sep 22, 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

1-2
Expected OA Rounds
52%
Grant Probability
73%
With Interview (+20.4%)
3y 9m (~1y 11m remaining)
Median Time to Grant
Low
PTA Risk
Based on 485 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