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 .
Response to Applicant’s Arguments
Applicant’s arguments filed March 10, 2026 have been fully considered, but they are deemed not persuasive.
Applicant argues that Ding’s weights in Blender does not teach the claimed “weight map that is generated based on at least one image included in some set of one or more images” as claimed. To a person of ordinary skill in the art, Ding’s Blender is a free, open-source 3D computer graphics suite handling everything from 3D modeling, sculpting, and rigging to animation, rendering, and video editing; especially, in Animation & Rigging, Ding’s Blend software had been known to build characters to life with skeletal structures, keyframing, and physics simulations (see Ding, 1.3. Thesis Objectives - In this study, we apply Blender, an open-source lightweight 3D animation software, to complete the dynamic transformation of the static model). Given a set of images, to a person of ordinary skill in the art at the time the invention was made, a model could be obviously built (Ding, 1.4. Thesis Contributions - Design and development of a system integrated from multiple processing stages to obtain 3D animated model from a single 2D image and a human motion video); Ding’s weights are generated from a conventional rigging process for creating and assigning weights to skeletons of the 3D animated model (e.g., Ding, Figure 2.4: Rigging Function); Ding’s weights assigned to skeleton’s joints (known as weight map – Ding, 5.1. Rigging Principle - Blender can connect bones and vertices by grouping points in a mesh. Generally, each bone can control a vertex group of its surrounding vertices. The same vertex can exist inside multiple vertex groups at the same time. In Blender, weights are used to indicate the amount of influence a bone has on a vertex) are based on the 3D animated model, therefore, also based on the captured 2D image of the character from which the 3D animated model was built (see also Ding, 5.4.1 Comparison with Other Models - The primary goal of our work is to quickly integrate BVH file and 3D humans reconstructed from a single photograph… The larger the number of vertices, the greater the workload of rigging and weight allocation). Accordingly, the claimed invention as represented does not represent a patentable distinction over the art of record.
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 1-4, 9-13, 17-18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over DING et al (3D Animation of a Human Body Reconstructed from a Single Photograph) in view of SCHMID et al (Programmable Motion Effects).
As per claim 1, Ding teaches the claimed "computer-implemented method for generating animations," the method comprising: "generating one or more images of an object based on user input" (Ding, 1. Image Acquisition Module, Figure 3.1: General System Architecture - As an alternative solution, this research chooses the method that reconstructs the 3D model of the human body from a single monocular photo for our case instead); "generating a weight map based on at least one image included in the one or more images" (Ding, 5.1. Rigging Principle - Blender can connect bones and vertices by grouping points in a mesh. Generally, each bone can control a vertex group of its surrounding vertices. The same vertex can exist inside multiple vertex groups at the same time. In Blender, weights are used to indicate the amount of influence a bone has on a vertex. The weight is between 0 and 1, indicating the degree of influence, with a weight of 1 indicating 100% influence and 0 indicating no influence); Ding suggests "textured geometry" of the claimed "generating textured geometry based on the one or more images" (Ding, 2.2.3. PIFu and PIFuHD - Real-life photographs of human bodies often depict them in varied settings, with various attire and hairstyles. Shunsuke Saito et al. developed the Pixel-aligned Implicit Function (PIFu) to digitize clothed humans and derive 3D surfaces and textures from a single picture. The PIFu is an Implicit Neural Representation (INR) capable of completely convolving to align the pixels of a 2D picture with the global context of the matching 3D object; 3.1. System Architecture Overview - Our system architecture is based on an integration of three primary approaches: To reconstruct human 3D models from a single picture using the PIFuHD pretrained model, to capture motions from sports videos using the MocapNET model, and to do the auto rigging and animation copy); see also Schmid, 3.1 Motion Effect Programs - It can evaluate the object's surface shaders as needed and draw upon additional scene information, such as the object's mesh data (vertex positions, normals, texture coordinates, etc.), auxiliary textures, the camera view vector, and vertex velocity vectors; 4.2 TAO Intersection - Each ray is intersected with the primitives of the TAO (mesh faces and bilinear patches) to compute all intersections along the ray, not just the one which is closest to the camera. Normals, texture coordinates, and other surface properties are interpolated linearly over the primitives to the intersection point and stored); and "generating an animation of the object based on the textured geometry, the weight map, and a skeleton" (Ding, 5.1.1. Kinematics in Animation - In Blender's action mode, it is possible to detect the action pattern of each body part under the influence of multiple controllers. In order to meet different movement needs, the weight and position of each part of the controller can be adjusted, and new controllers can be customized; 5.3. Animation Copy - After completing the rigging process, we get a 3D model with the controller. We can add motions to this 3D model. At this stage, we use motion replication to achieve motion synchronization between the 3D model and the skeleton animation from MocapNET).
Applicant’s arguments filed March 10, 2026 have been fully considered, but they are deemed not persuasive.
Applicant argues that Ding’s weights in Blender does not teach the claimed “weight map that is generated based on at least one image included in some set of one or more images” as claimed. To a person of ordinary skill in the art, Ding’s Blender is a free, open-source 3D computer graphics suite handling everything from 3D modeling, sculpting, and rigging to animation, rendering, and video editing; especially, in Animation & Rigging, Ding’s Blend software had been known to build characters to life with skeletal structures, keyframing, and physics simulations (see Ding, 1.3. Thesis Objectives - In this study, we apply Blender, an open-source lightweight 3D animation software, to complete the dynamic transformation of the static model). Given a set of images, to a person of ordinary skill in the art at the time the invention was made, a model could be obviously built (Ding, 1.4. Thesis Contributions - Design and development of a system integrated from multiple processing stages to obtain 3D animated model from a single 2D image and a human motion video); Ding’s weights are generated from a conventional rigging process for creating and assigning weights to skeletons of the 3D animated model (e.g., Ding, Figure 2.4: Rigging Function); Ding’s weights assigned to skeleton’s joints (known as weight map – Ding, 5.1. Rigging Principle - Blender can connect bones and vertices by grouping points in a mesh. Generally, each bone can control a vertex group of its surrounding vertices. The same vertex can exist inside multiple vertex groups at the same time. In Blender, weights are used to indicate the amount of influence a bone has on a vertex) are based on the 3D animated model, therefore, also based on the captured 2D image of the character from which the 3D animated model was built (see also Ding, 5.4.1 Comparison with Other Models - The primary goal of our work is to quickly integrate BVH file and 3D humans reconstructed from a single photograph… The larger the number of vertices, the greater the workload of rigging and weight allocation).
It would have been obvious, in view of Schmid, to configure Ding's method as claimed by generating textured geometry based on the inputted image(s). The motivation is to create the textured 3D object based on the 3D model generated from the weight set and the skeleton.
Claim 2 adds into claim 1 "determining one or more joint positions based on the at least one image; and generating the skeleton based on the one or more joint positions" (Ding, 5.2.1. Introduction to Three Methods of Rigging - The second is a paid plug-in for blender called Auto Rig Pro. for humanoid characters. Auto-Rig Pro is an all-in-one solution to rig characters, retarget animations, and provide 3D export for Unity and Unreal Engine. It provides functions that specify a few body joints to rig the skeleton automatically).
Claim 3 adds into claim 1 "wherein generating the one or more images comprises processing the user input via a trained machine learning model to generate the one or more images" (Ding, 4.1. PIFuHD Introduction - As illustrated in Figure 4.1, PIFuHD employs coarse-to-fine two-level pipeline for fine surface reconstruction. This research uses the pre-trained PIFuHD model to generate the result containing the desired 3D human body, which is the fundamental step of the animation).
Claim 4 adds into claim 3 "wherein the trained machine learning model is configured to generate the one or more images based on at least one of a predefined style embedding or a predefined pose" (Ding, As Figure 5.8 shows, the 3D human can move according to skeletal animation after rigging and animation copy; Figure 4.8: The Display of Skeletal Animation Saved as BVH file in Blender).
Claim 9 adds into claim 1 "wherein generating the textured geometry comprises projecting the one or more images onto either planar mesh geometry or three- dimensional (3D) geometry that is generated via a trained machine learning model based on the one or more images" (Ding, 2.2.3. PIFu and PIFuHD - Real-life photographs of human bodies often depict them in varied settings, with various attire and hairstyles. Shunsuke Saito et al. developed the Pixel-aligned Implicit Function (PIFu) to digitize clothed humans and derive 3D surfaces and textures from a single picture. The PIFu is an Implicit Neural Representation(INR) capable of completely convolving to align the pixels of a 2D picture with the global context of the matching 3D object; 3.1. System Architecture Overview - Our system architecture is based on an integration of three primary approaches: To reconstruct human 3D models from a single picture using the PIFuHD pretrained model, to capture motions from sports videos using the MocapNET model, and to do the auto rigging and animation copy; Schmid, 3.1 Motion Effect Programs - It can evaluate the object's surface shaders as needed and draw upon additional scene information, such as the object's mesh data (vertex positions, normals, texture coordinates, etc.), auxiliary textures, the camera view vector, and vertex velocity vectors; 4.2 TAO Intersection - Each ray is intersected with the primitives of the TAO (mesh faces and bilinear patches) to compute all intersections along the ray, not just the one which is closest to the camera. Normals, texture coordinates, and other surface properties are interpolated linearly over the primitives to the intersection point and stored). It would have been obvious, in view of Schmid, to configure Ding's method as claimed by projecting the one or more images onto either planar mesh geometry or three-dimensional (3D) geometry. The motivation is to create the textured 3D object based on the 3D model generated from the weight set and the skeleton.
Claim 10 adds into claim 1 "wherein the object is a character" (Ding, Figure 5.8: Dynamic 3D Human).
Claims 11-13, 18, 20 claim one or more non-transitory computer-readable storage media and a system based on the method of claims 1-4, 9-10; therefore, they are rejected under a similar rationale.
Claim 17 adds into claim 11 "wherein the one or more images include an image of a back of the object and an image of a front of the object" (Ding, Figure 4.1 - Front/back normal images; Figure 3.3 - natural pose).
Claims 5-7, 14-16 are rejected under 35 U.S.C. 103 as being unpatentable over DING et al (3D Animation of a Human Body Reconstructed from a Single Photograph) in view of SCHMID et al (Programmable Motion Effects), and further in view of BASSET et al (SMEAR: Stylized Motion Exaggeration with ARt-direction).
Claim 5 adds into claim 1 "wherein generating the weight map comprises: processing the at least one image via a trained machine learning model to determine a pose of the object" (Ding, 1.3. Thesis Objectives - The prerequisite for accomplishing pose sequence recognition is motion capture, measuring, tracking, and recording the motion trajectory of an object in 3D space. To meet this demand of our target system, we utilize the MocapNET which trains a feedforward neural network (FNN) that regresses directly from two-bit joint rotations; 4.1. PIFuHD Introduction - As illustrated in Figure 4.1, PIFuHD employs coarse-to-fine two-level pipeline for fine surface reconstruction. This research uses the pre-trained PIFuHD model to generate the result containing the desired 3D human body, which is the fundamental step of the animation); "generating a segmentation of the object based on the pose of the object" (Basset, 3.2 Articulated figure - Single Bone. The geometry for a single bone is illustrated in Figure 5 Mathematical notations for a single bone. For every vertex i of the mesh with position pi, we compute the normal n^(ui) of the ribbon (in green) which is orthogonal to the bone axis b = (ct - cr) and is as aligned as possible to the direction of motion ^v(u) interpolated from the normalized velocities v^r and v^t at root and tip locations cr and ct); and "performing one or more blurring operations across one or more edges of one or more segments included in the segmentation to generate the weight map" (Basset, Figure 11: Elongated in-betweens created with different stylization functions (figures 11(b)-11(d) - blurring left foot) for the middle frame of a character animation; see also Schmid, 5.2 Examples - Pinocchio's Peckish Pest The last set of examples show our results on a production-quality scene. Figure 10 uses a combination of speed lines and weighted motion blur. The motion blur's weighting curve has a spike around the current time, so that Pinocchio is shown clearly in every frame, even when he is moving quickly. In Figure 11 a comparison between conventional motion blur (also rendered by our system) and our weighted motion blur is made; Figure 10: Results from applying speed line and motion blur effects (bottom), and the stroboscopic image effect (top) to an animation of Pinocchio and a woodpecker). It would have been obvious, in view of Basset and Schmid, to configure Ding's method as claimed by performing one or more blurring operations across one or more edges of one or more segments. The motivation is to form the smearing/blurring effect on the character's motion.
Claim 6 adds into claim 5 "wherein performing the one or more blurring operations comprises: determining the one or more edges of the one or more segments" (Basset, 3.2 Articulated figure - Single Bone. The geometry for a single bone is illustrated in Figure 5... Mathematical notations for a single bone. For every vertex i of the mesh with position pi, we compute the normal n^(ui) of the ribbon (in green) which is orthogonal to the bone axis b = (ct - cr) and is as aligned as possible to the direction of motion v^(u) interpolated from the normalized velocities v^r and v^t at root and tip locations cr and Ct; Figure 8: Motion offsets computed for the animation in (a) with different normalization factors (b-c). In (d), we compare the motion offsets computed considering all bones (left) with those obtained by considering the hand as a single body part affected by the forearm bone (right)); and "computing, for each segment included in the one or more segments, a corresponding direction based on a plurality of joints associated with the segment" (Basset, Figure 16: The stylization effects achievable by our method may be combined to create complex motion stylizations). It would have been obvious, in view of Basset and Schmid, to configure Ding's method as claimed by performing one or more blurring operations across one or more edges of one or more segments. The motivation is to form the smearing/blurring effect on the character's motion.
Claim 7 adds into claim 1 "wherein generating the weight map comprises: determining a longest axis associated with the object in the at least one image; segmenting the object into a plurality of segments along the longest axis" (Ding, 2.4. Work on Rigging - As Figure 2.4 shows, rigging (also known as skeleton binding) process is setting up the skeleton-based animation to combine the 3D model and skeleton animation, to generate animated 3D animation) (it is noted that the selection of the "longest axis" is just an optional choice of axis of multiple segments in which each segment (e.g., Figure 5.4: Metarig and Rig with Controllers in Rigify) can be controlled for motioning); and "performing one or more blurring operations across one or more edges of the plurality of segments to generate the weight map" (Basset, Figure 11: Elongated in-betweens created with different stylization functions (figures 11(b)-11(d) - blurring left foot) for the middle frame of a character animation; see also Schmid, 5.2 Examples - Pinocchio's Peckish Pest The last set of examples show our results on a production-quality scene. Figure 10 uses a combination of speed lines and weighted motion blur. The motion blur's weighting curve has a spike around the current time, SO that Pinocchio is shown clearly in every frame, even when he is moving quickly. In Figure 11 a comparison between conventional motion blur (also rendered by our system) and our weighted motion blur is made; Figure 10: Results from applying speed line and motion blur effects (bottom), and the stroboscopic image effect (top) to an animation of Pinocchio and a woodpecker). It would have been obvious, in view of Basset and Schmid, to configure Ding's method as claimed by segmenting the character into different sections and performing the blurring effect on the segmented section. The motivation is to form the smearing/blurring effect on the character's motion.
Claims 14-16, claim one or more non-transitory computer-readable storage media based on the method of claims 5-7, and 9; therefore, they are rejected under a similar rationale.
Claims 8, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over DING et al (3D Animation of a Human Body Reconstructed from a Single Photograph) in view of REN et al (Make-A-Character: High Quality Text-to-3D Character Generation within Minutes) and MAHAJAN et al (Prompting Hard or Hardly Prompting: Prompt Inversion for Text-to-Image Diffusion Models)
In claim 8, Ding does not teach, but Ren and Mahajan teach "the user input comprises text describing at least one aspect of the object" (Mahajan, 3. Prompt Inversion for Diffusion, Figure 3. Effects of Timesteps on Prompt Inversion. Prompts generated for the different sections of timesteps in text-to-image diffusion models with maximum noise at timestep T. The images generated show that the semantic information is completely recovered for the range starting from the middle timesteps. Once the semantic information appears in the images, only image-level details change when considering the larger range starting from the early steps of the forward diffusion process; Ren, Figure 1 - The Mach system is capable of creating highly detailed and varied 3D avatars from text prompts. We demonstrate the versatility of these characters by showcasing their ability to express dynamic animations through various facial expressions). Thus, it would have been obvious, in view of Mahajan and Ren, to configure Ding's method as claimed by generating one or more prompts based on inputting the one or more prompts into a trained diffusion model that generates the one or more images. The motivation is to allow users to construct (and edit) effective textual descriptions of concepts they desire.
Claim 19 claims one or more non-transitory computer-readable storage media based on the method of claim 8; therefore, it is rejected under a similar rationale.
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PHU K NGUYEN whose telephone number is (571)272-7645. The examiner can normally be reached M-F 8-5pm.
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, Daniel F. Hajnik 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.
/PHU K NGUYEN/Primary Examiner, Art Unit 2616
Prosecution Insights