DETAILED ACTION
Status
This Office Action is responsive to claims filed on 03/30/2025. Please note Claims 1-20 are pending and have been examined.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
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 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-4, 7-9, 17, 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Zheng (US 20260237132 A1), in view of Chung (“LucidDreamer: Domain-Free Generation of 3D Gaussian Splatting Scenes”, cited on IDS 7/14/2025).
Regarding Claim 1, Zheng discloses a method comprising:
receiving, by a processing device, a text-based query that specifies a (Fig. 1, 116. [0036] “The first stage generates a 3D representation of a scene or object described by the text prompt 116.”);
generating, by processing the text-based query using one or more machine learning models (Fig. 1, 102 “PRE-TRAINED 3D DIFFUSION MODEL”), a (Fig. 1, 108. [0036] “The first stage generates a 3D representation of a scene or object described by the text prompt 116. The 3D representation is embodied in a 3D neural radiance field 106 volume.”); and
constructing, for output (Fig. 1, 120 “RENDERER”) by the processing device, a three-dimensional environment that includes a traversable ([0036] “…to generate movement (coordinate changes) in three dimensions”) three-dimensional scene that depicts the visual aspect, the constructing including processing the spherical representation using the one or more machine learning models (Fig. 1, 114. [0036] “FIG. 2 depicts an example of a 3D neural radiance field 106 in multiple feature resolutions processed through the motion rendering model's 4D scene generating model 114 to generate movement (coordinate changes) in three dimensions.”).
Zheng does not expressly disclose the text-based query that specifies a visual aspect to be included in a three-dimensional scene, and the spherical representation of the three-dimensional scene.
However, in the same field of endeavor, Chung discloses receiving a text-based query that specifies a visual aspect to be included in a three-dimensional scene (see text prompt in Fig. 2 or Fig. 3, e.g. “a cat on the street near a white house with stairs, trees nearby, and gray blocks”) and generating a spherical representation of the three-dimensional scene (see Section II RELATED WORK, A. 3D Scene Representation, “In particular, 3D Gaussian splatting [50] utilizes the concept of Gaussian splats combined with spherical harmonics and opacity to represent complete and unbounded 3D scenes.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to have modified the method of Zheng with the feature of receiving a text-based query that specifies a visual aspect and generating a spherical representation of the three-dimensional scene. Doing so could “synthesize diverse, high-quality, and expandable 3D scenes”, as taught by Chung.
Regarding Claim 2, Zheng-Chung discloses the method as described in claim 1, wherein the traversable three-dimensional scene includes a 3D Gaussian Splat representation (see Chung Section II RELATED WORK, A. 3D Scene Representation, “In particular, 3D Gaussian splatting [50] utilizes the concept of Gaussian splats combined with spherical harmonics and opacity to represent complete and unbounded 3D scenes.”).
Regarding Claim 3, Zheng-Chung discloses the method as described in claim 1, wherein the generating the spherical representation includes using a text-to-video diffusion model fine-tuned on panoramic video data to process the text-based query (see Chung, first page, right column, “Studies such as RGBD2 [12] and BlockFusion [13], which train diffusion models on limited datasets [14], [15] for 3D scene synthesis… Another prominent approach is directly generating 3D scenes using a 2D diffusion model combined with a depth estimation model.”).
Regarding Claim 4, Zheng-Chung discloses the method as described in claim 1, wherein the spherical representation includes a sequence of frames of a 360-degree panoramic video (Chung, first page, right column, “DreamScene360 [26] generates a panoramic image and constructs a 3D scene using estimated depth.”).
Regarding Claim 7, Zheng-Chung discloses the method as described in claim 1, further comprising training the one or more machine learning models using a combination of panoramic video data and panoramic image data as training data (Zheng [0025] “The disclosed mechanisms utilize a combination of 2D, 3D, and video-trained diffusion models to disengage the 3D geometry expressed in the text prompt from motion expressed in the text prompt.” Chung Section II RELATED WORK, B. 3D Scene Generation, “These models can generate high-quality 3D content due to the power of diffusion models; however, the results show low diversity, as the 3D dataset used for training the diffusion models is scarcer than the 2D dataset. One approach to overcoming the scarcity of 3D datasets is to utilize pre-trained 2D diffusion models to lift images into 3D space.”).
Regarding Claim 8, Zheng-Chung discloses the method as described in claim 7, wherein the training further includes applying a masked diffusion loss to exclude one or more regions of individual panoramic videos from the training data (Chung Section III METHOD, A. Point Cloud Construction, “We define the mask Mi to discriminate the region that is filled by existing points in ˆ Ii. Specifically, the Mi value is set to one if the corresponding pixel is already filled or zero otherwise”).
Regarding Claim 9, Zheng-Chung discloses the method as described in claim 1, further comprising rendering views of the traversable three-dimensional scene not represented by the spherical representation based on an input to navigate within the three-dimensional environment (Chung, second page, left column, “In the Navigation stage, the camera moves along a predefined trajectory, and the visible region of the point cloud is projected onto the new camera plane.” V CONCLUSION, “We introduce a novel pipeline, LucidDreamer, which harnesses the power of a state-of-the-art image diffusion model to generate 3D scenes across diverse domains. Our pipeline operates iteratively through three stages: Navigation, Dreaming, and Alignment. In the Navigation stage, the camera is directed toward the area where the scene needs to be expanded.”).
Regarding Claim 17, it recites similar limitations of claim 1. The rationale of claim 1 rejection is applied to reject claim 17. (Notice that the “outputting” step is taught by Zheng Fig. 1, 120 and [0074])
Regarding Claim 19, it recites similar limitations of claim 9. See claim 9 rejection above for mappings.
Regarding Claim 20, it recites similar limitations of claim 3. See claim 3 rejection above for mappings.
Claims 5-6, 10-16 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Zheng in view of Chung, further in view of ANONYMOUS AUTHOR(S), “Text-based Fully Traversable 3D Scene Generation from Panoramas”, cited on IDS 7/14/2025 (hereinafter, “3DSG”).
Regarding Claim 5, Zheng-Chung discloses the method as described in claim 4. In the same field of endeavor, 3DSG discloses wherein the constructing the three-dimensional environment includes: extracting one or more perspective views from each frame of the sequence of frames (Section 3.3, “For each panoramic frame, we randomly perform 3 perspective crops with a field of view of 120°, resulting in 384 perspective images for reconstruction.”); determining virtual poses for each of the extracted perspective views (Section 3.3, “Similar to standard scene reconstruction, we calibrate the camera poses of these perspective images using COLMAP [Schonberger and Frahm 2016].”); and processing the one or more perspective views and the virtual poses using a three-dimensional reconstruction model (Section 3.3, “We adopt long-LRM, the state-of-the-art feed-forward 3DGS reconstructor [Ziwen et al. 2024a] to reconstruct the final 3D scene using the estimated camera poses and generated views.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to have modified the method of Zheng-Chung with the feature of extracting one or more perspective views from each frame of the sequence of frames, determining virtual poses for each of the extracted perspective views, and processing the one or more perspective views and the virtual poses using a three-dimensional reconstruction model. Doing so could “achieve fast reconstruction, requiring a processing time of only 1.3 seconds in total”, as taught by 3DSG.
Regarding Claim 6, Zheng-Chung-3DSG discloses the method of claim 5, wherein the three-dimensional reconstruction model is a long-LRM model (3DSG, Section 3.3, “We adopt long-LRM, the state-of-the-art feed-forward 3DGS reconstructor [Ziwen et al. 2024a] to reconstruct the final 3D scene using the estimated camera poses and generated views.”).
Regarding Claim 10, it recites similar limitations of claim 5. The rationale of claim 5 rejection is applied to reject claim 10.
Regarding Claim 11, it recites similar limitations of claim 3. See claim 3 rejection above for mappings.
Regarding Claim 12, it recites similar limitations of claim 8. See claim 8 rejection above for mappings.
Regarding Claim 13, it recites similar limitations of claim 6. See claim 6 rejection above for mappings.
Regarding Claim 14, it recites similar limitations of claim 7. See claim 7 rejection above for mappings.
Regarding Claim 15, it recites similar limitations of claim 9. See claim 9 rejection above for mappings.
Regarding Claim 16, it recites similar limitations of claim 2. See claim 2 rejection above for mappings.
Regarding Claim 18, it recites similar limitations of claim 5. See claim 5 rejection above for mappings.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Tatarchenko (US 20260024273 A1) – This reference discloses a device and a computer implemented method for generating a synthetic digital image of a three-dimensional scene, in particular for a dataset for training and/or testing of a machine learning system. The method includes providing at least one text prompt which includes a description of a three-dimensional layout of the scene, wherein the at least one text prompt comprises a description of a style of the scene, generating the layout depending on the description of the layout, assembling the scene depending on the layout, determining a three-dimensional Gaussian Splatting representation of the assembled scene depending on the assembled scene, rendering a digital image from the three-dimensional Gaussian Splatting representation, and determining the synthetic digital image with a stable diffusion depending on the digital image and the description of the style.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHONG WU whose telephone number is (571)270-5207. The examiner can normally be reached MON-FRI: 9AM-5PM 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.
/CHONG WU/Primary Examiner, Art Unit 2613
Prosecution Insights