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 .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 04/27/2026 has been entered.
Response to Arguments
Applicant's arguments filed 04/27/2026 have been fully considered but they are not persuasive.
The Applicant alleged the following: “The Examiner's interpretation of Ramirez de Chanlatte's "without point clouds "disclosure as the same as an "incomplete point-cloud" is unreasonable in light of Applicant’s disclosure and the amended claims. Applicant's Specification uses "incomplete point cloud" to mean a partial-but-present point cloud whose missing portions arise from limited viewpoints, occlusion, or low-resolution sampling, not the absence of any point-cloud input. "The incomplete point cloud comprises a set of 3D input points P captured by a depth sensor or LiDAR sensor." See Specification at paragraphs [0002] and [0024]. Under the broadest reasonable interpretation, "an incomplete point-cloud comprising one or more points" cannot reasonably be read as the absence of a point-cloud input. Thus, Ramirez de Chanlatte teaches away from the amended requirement of receiving an incomplete point-cloud comprising one or more points captured by a sensor and using that partial point-cloud in reconstructing the 3D surface. Because Ramirez de Chanlatte teaches away from the claimed invention, one of ordinary skill in the art would not modify Ramirez de Chanlatte to incorporate certain features of the combination of Achlioptas, Larson, and Kavanau in an effort to arrive at the claimed invention. Furthermore, the Office Action fails to provide the articulated reason with rational underpinning required by KSR for modifying the depth-map-based system disclosed by Ramirez de Chanlatte, which expressly avoids point clouds, into Applicant's claimed text-guided point-cloud completion framework. Accordingly, Applicant respectfully submits that the rejection is improper and respectfully requests that the rejection of claims 1-3, 5-7, and 9-22 be withdrawn.” The examiner is not persuaded. The examiner is not persuaded. In this case, we find such a modification of predicted noise to be obvious. In KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 417 (2007), the Supreme Court held that “if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious unless its actual application is beyond his or her skill.” Id. at 417. “The combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results.” Id. at 416; see also id. at 417 (“If a person of ordinary skill can implement a predictable variation, § 103 likely bars its patentability.”); In re Schreiber, 128 F.3d 1473, 1477 (Fed. Cir. 1997) (“It is well settled that the recitation of a new intended use for an old product does not make a claim to that old product patentable.” (citations omitted)). We do not find that the evidence shows having predicted noise “uniquely challenging or difficult for one of ordinary skill in the art.” Leapfrog Enters., Inc. v. Fisher-Price, Inc., 485 F.3d 1157, 1162 (Fed. Cir. 2007) (citing KSR, 550 U.S. at 418). Accordingly, we do not consider the Applicant’s argument to sufficiently demonstrate the Examiner’s rejection is in error. Accordingly, the examiner maintains the rejection.
The Applicant alleged the following: “Additionally, the cited references fail to teach the integrated claim requirement of receiving an incomplete point-cloud for the object captured by a sensor at a position, processing a text description associated with the object and a rendered image of the 3D surface of the object with noise by a text-to-image model to predict the noise, and updating the 3D surface based on the predicted noise and the incomplete point-cloud.” The examiner is not persuaded. As mentioned previously, the examiner interprets paragraph 0004 teachings of “conventional systems often require commonly unavailable inputs include point clouds” as being the same as “incomplete point cloud” data. Moreover, Ramirez de Chanlatte discloses the Applicant’s claim language of “captured by a sensor at a position” in Paragraphs 0070; 0107; 0133. Ramirez de Chanlatte goes on to discloses “processing a text description associated with the object” in Paragraphs 0114; 0124-0129. Ramirez de Chanlatte Paragraphs 0033; 0124-0125 discloses the Applicant’s claim language of “and a rendered image of the 3D surface of the object.” Larson Paragraph 0070 discloses the Applicant’s claim language of “with noise by a text-to-image model.” Achlioptas Paragraphs 0040; 0061 discloses predict the noise. Accordingly, the examiner maintains the rejection.
The Applicant alleged the following: “Ramirez de Chanlatte does not teach processing a text description together with a rendered image of the 3D surface by a text-to-image model, much less updating the 3D surface by reconstructing a missing portion according to the text description. At paragraph [0114] Ramirez de Chanlatte discloses user-interface editing operations such as inserting text onto an object surface, and paragraph [0124] discloses presenting text and rendered reconstructed 3D models in a user interface.” The examiner is not persuaded. The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses the Applicant’s claim language. Moreover, Larson discloses using a text-to-image model in Paragraph 0070. Accordingly, the examiner maintains the rejection.
The Applicant alleged the following: “Achlioptas and Larson do not remedy the deficiencies of Ramirez de Chanlatte. Achlioptas discloses denoising diffusion and predicted noise in a general diffusion framework and discloses noisy point clouds in the context of garment generation from scribbles, but Achlioptas does not disclose processing a text description associated with the object together with a noisy rendered image of the current 3D surface by a text-to-image model. See Achlioptas at paragraphs [0040] and [0061].” The examiner is not persuaded. Ramirez de Chanlatte goes on to discloses “processing a text description associated with the object” in Paragraphs 0114; 0124-0129. Ramirez de Chanlatte Paragraphs 0033; 0124-0125 discloses the Applicant’s claim language of “and a rendered image of the 3D surface of the object.” Larson Paragraph 0070 discloses the Applicant’s claim language of “with noise by a text-to-image model.” Achlioptas Paragraphs 0040; 0061 discloses predict the noise. Accordingly, the examiner maintains the rejection.
The Applicant alleged the following: “The Office Action does not provide the articulated reason with rational underpinning required by KSR and MPEP § 2143 for combining Ramirez de Chanlatte, Achlioptas, and Larson in the manner asserted. The rejection thus depends on impermissible gap-filling: Ramirez de Chanlatte does not use point clouds, Achlioptas does not teach the claimed text-conditioned noisy-render processing, and Larson does not disclose using a text-to-image model in the claimed reconstruction/update workflow.” The examiner is not persuaded.
Ramirez de Chanlatte indeed discloses point clouds in Paragraphs 0004; 0036; 0062. In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., text-conditioned noisy-render processing) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). Additionally, Larson discloses using a text-to-image model in Paragraph 0070. Accordingly, the examiner maintains the rejection.
The Applicant alleged the following: “Furthermore, Kavanau also fails to cure the foregoing deficiencies. Kavanau discloses increasing a probability of deviation of an azimuth or random initialization. The cited combination still does not teach or suggest receiving and using an incomplete point-cloud comprising one or more points for the object captured by a sensor at a position, or processing a text description associated with the object together with a rendered image of the 3D surface of the object with noise by a text-to-image model to predict the noise.” The examiner is not persuaded.
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses the Applicant’s claim language. Moreover, the combination of Ramirez de Chanlatte, Achlioptas and Larson discloses the following:
- receiving an incomplete point-cloud comprising one or more points for the object (See Ramirez de Chanlatte Paragraphs 0004; 0036; 00621)
-captured by a sensor at a position (See Ramirez de Chanlatte Paragraphs 0070; 0107; 0133);
-processing a text description associated with the object (See Ramirez de Chanlatte Paragraphs 0114; 0124-0129)
and a rendered image of the 3D surface of the object (See Ramirez de Chanlatte Paragraphs 0033; 0124-0125)
-with noise by a text-to-image model (See Larson Paragraph 00702)
-to predict the noise (See Achlioptas Paragraphs 0040; 0061);
Accordingly, the examiner maintains the rejection.
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.
Claim(s) 1-3, 5, 7, 9-20 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Ramirez de Chanlatte, US 20230147722 and in view of Achlioptas, US 20240112401.
Claim 1:
Ramirez de Chanlatte discloses a computer-implemented method for reconstructing a three-dimensional (3D) surface model of an object (See Ramirez de Chanlatte Abstract) but fails to explicitly disclose predicted noise. Achlioptas discloses this feature in paragraphs 0040; 0061. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have further modified Ramirez de Chanlatte by the teachings of Achlioptas to enable improved generation of three-dimensional (3D) objects by prediction of noise, more effectively. In addition, both references teach features that are directed to analogous art and they are directed to the same field on endeavor, such as generating (reconstruction) of 3D images using point clouds.
Moreover, the combination of Ramirez de Chanlatte and Achlioptas failed to disclose a text-to-image model. However, Larson discloses this feature in Paragraph 0070. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have further modified Ramirez de Chanlatte and Achlioptas by the teachings of Larson to enable improved machine learning training data for improving a predictive accuracy of a machine learning models, more effectively (See Larson Abstract).
As modified:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses the following:
initializing the 3D surface of the object (See Ramirez de Chanlatte Figure 5 and Paragraphs 0006; 0033; 0089);
receiving an incomplete point-cloud comprising one or more points for the object (See Ramirez de Chanlatte Paragraphs 0004; 0036; 00623) captured by a sensor at a position (See Ramirez de Chanlatte Paragraphs 0070; 0107; 0133);
processing a text description associated with the object (See Ramirez de Chanlatte Paragraphs 0114; 0124-0129) and a rendered image of the 3D surface of the object (See Ramirez de Chanlatte Paragraphs 0033; 0124-0125) with noise by a text-to-image model (See Larson Paragraph 00704) to predict the noise (See Achlioptas Paragraphs 0040; 0061); and
updating, based on the predicted noise (See Achlioptas Paragraph 0040), at least a portion of the 3D surface of the object (See Ramirez de Chanlatte Paragraphs 0005-0006; 0017-0022; 0033; 0089) by reconstructing a missing portion of the 3D surface ((See Ramirez de Chanlatte Paragraph 00055) according to the text description (See Ramirez de Chanlatte Paragraph 0124).
Claim 2:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein the 3D surface (See Ramirez de Chanlatte Paragraphs 0006; 0033; 0089) comprise at least one of a signed distance function (See Ramirez de Chanlatte Paragraphs 0006; 0036), polygonal mesh (See Ramirez de Chanlatte Paragraph 0034), or a neural surface (See Ramirez de Chanlatte Paragraph 0017-0018; 0033-0039) and a volumetric coloring function (See Ramirez de Chanlatte Paragraphs 0045; 0111; 0114; 0115).
Claim 3:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein updating the 3D surface of the object (See Ramirez de Chanlatte Figure 5 and Paragraphs 0006; 0033; 0089; 0100-0102) reduces one or more differences between the noise and the predicted noise (See Achlioptas Paragraphs 0040; 0061) and reduces one or more differences between the incomplete point-cloud (See Ramirez de Chanlatte Paragraphs 0004; 0062) and the 3D surface of the object (See Ramirez de Chanlatte Figure 5 and Paragraphs 0006; 0033; 0089; 0100-0102).
Claim 5:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses
rendering additional images of the 3D surface of the object (See Ramirez de Chanlatte Figure 5 and Paragraphs 0006; 0089) according to different camera viewpoints (See Ramirez de Chanlatte Paragraphs 00326; 0042; 0071);
combining sampled noise with the additional images to produce additional noisy images (See Achlioptas Paragraphs 0040; 0061);
updating the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033) to reduce one or more differences between the predicted noise and the sampled noise (See Achlioptas Paragraphs 0040; 0061).
Claim 7:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein the different camera viewpoints (See Ramirez de Chanlatte Paragraphs 00327; 0042; 0071) are associated with natural poses (See Ramirez de Chanlatte Paragraph 0042) of the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033).
Claim 9:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein updating the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033) encourages locations on the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033) to go through input points of the incomplete point-cloud (See Ramirez de Chanlatte Paragraphs 0004; 0062).
Claim 10:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein updating the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033) discourages (reduces) locations on the 3D surface if the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033) between the position of the sensor and the incomplete point-cloud (See Ramirez de Chanlatte Paragraphs 0004; 0062).
Claim 11:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein updating the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022) discourages locations on the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 033) in an empty space outside a visual cone of the incomplete point-cloud (See Ramirez de Chanlatte Paragraphs 0004; 0062).
Claim 12:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein at least one of the steps of receiving, processing, or updating is performed on a server (See Ramirez de Chanlatte Paragraphs 0161-0163) or in a data center, and the 3D surface of the object (See Ramirez de Chanlatte Figure 6; Paragraphs 0005-0006; 0017-0022; 0033) is streamed to a user device (See Ramirez de Chanlatte Paragraphs 0050-0054).
Claim 13:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein at least one of the steps of receiving, processing, or updating is performed within a cloud computing environment (See Ramirez de Chanlatte Paragraphs 0050-0054).
Claim 14:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein at least one of the steps of receiving, processing, or updating is performed for training, testing, or certifying a neural network employed in a machine, robot, or autonomous vehicle (See Ramirez de Chanlatte Paragraphs 0017; 0064).
Claim 15:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein at least one of the steps of receiving, processing, or updating is performed on a virtual machine comprising a portion of a graphics processing unit (See Achlioptas Paragraphs 0040; 0076).
Claims 16-18:
Claims 16-18 are rejected on the same basis as claims 1-3.
Claims 19 and 20:
Claims 19 and 20 are rejected on the same basis as claims 1 and 2.
Claim 22:
The combination of Ramirez de Chanlatte, Achlioptas and Larson discloses wherein the text-to-image model (See Larson Paragraph 00708) comprises a pre-trained text-to-image model (See Larson Paragraph 00709) that has learned a semantic (See Achlioptas Paragraph 0039) prior associated with shapes of objects by synthesizing images (See Ramirez de Chanlatte Paragraphs 0017; 0021-0023) corresponding to text description inputs (See Ramirez de Chanlatte Paragraphs 0114; 0124-0129) and noisy images (See Achlioptas Paragraphs 0040; 0061).
Claims 6 and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Ramirez de Chanlatte, US 20230147722, in view of Achlioptas, US 20240112401, and in further view of Kavanau, US 20090274375.
Claim 6:
The combination of Ramirez de Chanlatte, Achlioptas and Larson failed to disclose increase a probability of deviation of an azimuth. However, Kavanau discloses this feature in paragraph 0048. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have further modified the combination of Ramirez de Chanlatte, Achlioptas and Larson by the teachings of Kavanau to enable improved 3D reconstruction by incorporating the increase of a probability of deviation of an azimuth, more effectively. In addition, both references teach features that are directed to analogous art and they are directed to the same field on endeavor, such as 3D reconstruction.
As modified:
The combination of Ramirez de Chanlatte, Achlioptas, Larson and Kavanau discloses the following:
wherein the different camera viewpoints (See Ramirez de Chanlatte Paragraphs 003210; 0042; 0071) progressively increase a probability of deviation of an azimuth compared with the position of the sensor (See Kavanau Paragraph 0048).
Claim 21:
The combination of Ramirez de Chanlatte, Achlioptas, Larson and Kavanau discloses wherein the 3D surface is initialized (See Kavanau Paragraph 0011) or to define either a sphere (See Ramirez de Chanlatte Paragraph 0109) or a 3D shape based on the incomplete point-cloud (See Ramirez de Chanlatte Paragraphs 0004; 006211).
Contact Information
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHEREE N BROWN whose telephone number is (571)272-4229. The examiner can normally be reached M-F 5:30-2:00 PM 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, SAID BROOME can be reached at (571) 272-2931. 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.
/SHEREE N BROWN/Primary Examiner, Art Unit 2612 May 15, 2026
1 The examiner interprets paragraph 0004 teachings of “conventional systems often require commonly unavailable inputs include point clouds” as being the same as “incomplete point cloud” data.
2 Paragraph 0070 of Larson discloses “a document-image generator that interface with one or more generative deep learning models, such as generative adversarial networks or a text-to-image model (e.g., DALL-E and/or DALL-E 2, a transformer language model derived from GPT-3) for producing one or more corpora of document samples or image samples.”
3 The examiner interprets paragraph 0004 teachings of “conventional systems often require commonly unavailable inputs include point clouds” as being the same as “incomplete point cloud” data.
4 Paragraph 0070 of Larson discloses “a document-image generator that interface with one or more generative deep learning models, such as generative adversarial networks or a text-to-image model (e.g., DALL-E and/or DALL-E 2, a transformer language model derived from GPT-3) for producing one or more corpora of document samples or image samples.”
5 Ramirez de Chanlatte teaching in Paragraph 0005 of “generate reconstructed 3D models that better conform to the shape of digital objects in real images than existing systems and use such reconstructed 3D models to generate more realistic looking visual effects (e.g., shadows, relighting)” correlates to the Applicant’s claim language of “missing portion.”
6 Ramirez de Chanlatte, Paragraph 0032 specifically recites “a depth map can include depth information derived from LIDAR (light detection and ranging) images, stereo images, or multiple images from different viewpoints, etc.”
7 Ramirez de Chanlatte, Paragraph 0032 specifically recites “a depth map can include depth information derived from LIDAR (light detection and ranging) images, stereo images, or multiple images from different viewpoints, etc.”
8 Paragraph 0070 of Larson discloses “a document-image generator that interface with one or more generative deep learning models, such as generative adversarial networks or a text-to-image model (e.g., DALL-E and/or DALL-E 2, a transformer language model derived from GPT-3) for producing one or more corpora of document samples or image samples.”
9 Paragraph 0070 of Larson discloses “a document-image generator that interface with one or more generative deep learning models, such as generative adversarial networks or a text-to-image model (e.g., DALL-E and/or DALL-E 2, a transformer language model derived from GPT-3) for producing one or more corpora of document samples or image samples.”
10 Ramirez de Chanlatte, Paragraph 0032 specifically recites “a depth map can include depth information derived from LIDAR (light detection and ranging) images, stereo images, or multiple images from different viewpoints, etc.”
11 The examiner interprets paragraph 0004 teachings of “conventional systems often require commonly unavailable inputs include point clouds” as being the same as “incomplete point cloud” data.
Prosecution Insights