DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on/after Mar. 16, 2013, is being examined under the first inventor to file provisions of the AIA .
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.
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)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1, 3-6, 11-15, and 17-20 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Saito et al. ("PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization", published June 2020, 'SAITO-2020').
Regarding claim 1, SAITO-2020 discloses a method performed by … computer(s), the method comprising:
PNG
media_image1.png
585
740
media_image1.png
Greyscale
obtaining a conditioning image of an object (SAITO-2020; FIG. 1, leftmost highlighted sub-image; p. 81, right col.; “The goal of this work is to achieve high-fidelity 3d reconstruction of clothed humans from a single image at a resolution sufficient to recover detailed information such as fingers, facial features and clothing folds …” FIG. 2, upper-left ‘Input’; p. 83, right col.; “We present a multi-level approach towards higher fidelity 3D human digitization that takes 1024×1024 resolution images as input.”);
PNG
media_image2.png
555
1533
media_image2.png
Greyscale
initializing an observation set characterizing a surface in [3-D] space that represents a [3-D] model of the object (SAITO-2020; FIG. 2; p. 83, left col.; “To further improve the quality and fidelity of the reconstruction, we first predict normal maps for the front and back sides [observation set characterizing a surface in [3-D] space] in image space, and feed these to the network as additional input.” p. 84, right col.; “We found that if we instead shift part of this inference problem into the feature extraction stage, the network can produce sharper reconstructed geometry. To do this, we predict normal maps as a proxy for 3-D geometry in image space and provide these normal maps as features to the pixel-aligned predictors. The 3-D reconstruction is then guided by these maps to infer a particular 3D geometry [represents a [3-D] model of the object], making it easier for the MLPs to produce details. We predict the backside and frontal normals in image space using a pix2pixHD network, mapping from RGB color to normal maps. … we find that this produces plausible outputs for the unseen backside for sufficiently constrained problem domains, such as clothed humans.”);
updating the observation set to generate a final observation set, the updating comprising, at each of [several] sampling iterations (SAITO-2020; § 3.1; p. 83, right col.; “A large scale dataset synthetically generated by rendering hundreds of high-quality scanned 3D human mesh models is used to train the function f in an end-to-end fashion. Unlike voxel-based methods, PIFu does not produce a discretized volume as output, so training can be performed by sampling 3-D points and computing the occupancy loss at the sampled locations, without generating 3-D meshes. During inference, 3-D space is uniformly sampled to infer the occupancy, and the final iso-surface is extracted [generate a final observation set] with a threshold of 0.5 using marching cubes”);
generating, from the observation set as of the sampling iteration, an initial updated observation set (SAITO-2020; FIG. 2; pp. 83-84; “Our method is composed of two levels of PIFu modules: … (2) a fine level that focuses on adding more subtle details by taking the original 1024×1024 resolution image as input, and producing backbone image features of 512×512 resolution … Notably, the fine level module takes 3D embedding features extracted from the coarse level instead of the absolute depth value.” [The Examiner notes that FIG. 2 of SAITO-2020 depicts the front-and-back images downsampled at half resolution, yielding an initial updated observation set.]);
generating, from the initial updated observation set, features of the [3-D] model of the object (SAITO-2020; FIG. 2; pp. 83-84; “Our method is composed of two levels of PIFu modules: (1) a coarse level similar to PIFu, focusing on integrating global geometric information by taking the downsampled 512 × 512 image as input, and producing backbone image features of 128 × 128 resolution …” [The Examiner notes that SAITO-2020 teaches backbone image features of 128 × 128 resolution, which are analogous to a 128x128 coarse pixel-aligned implicit function, as well as PIFu, and 512x512 fine PIFu features.]); and
updating the observation set using the features of the [3-D] model of the object (SAITO-2020; p. 84, right col.; “… if we … shift part of this inference problem into the feature extraction stage, the network can produce sharper reconstructed geometry. To do this, we predict normal maps as a proxy for 3-D geometry in image space and provide these normal maps as features to the pixel-aligned predictors. The 3D reconstruction is then guided by these maps to infer a particular 3-D geometry, making it easier for the MLPs to produce details. We predict the backside and frontal normals in image space using a pix2pixHD network, mapping from RGB color to normal maps. … we find that this produces plausible outputs for the unseen backside for sufficiently constrained problem domains, such as clothed humans.”); and
generating a [3-D] model of the object from the final observation set (SAITO-2020; FIG. 2, rightmost sub-image ‘Reconstruction’; p. 84; § 3.3; “We found that if we instead shift part of this inference problem into the feature extraction stage, the network can produce sharper reconstructed geometry. To do this, we predict normal maps as a proxy for 3D geometry in image space and provide these normal maps as features to the pixel-aligned predictors. The 3-D reconstruction is then guided by these maps to infer a particular 3-D geometry, making it easier for the MLPs to produce details. We predict the backside and frontal normals in image space using a pix2pixHD network, mapping from RGB color to normal maps. … we find that this produces plausible outputs for the unseen backside for sufficiently constrained problem domains, such as clothed humans.”).
Regarding claim 14, SAITO-2020 discloses … non-transitory computer-readable storage media encoded with instructions (SAITO-2020; p. 82, left col.; “We base our method on the recently introduced Pixel-Aligned Implicit Function (PIFu) representation.” [The Examiner asserts that PIFu is an algorithmic method, which is analogous to ‘instructions’.]) that, when executed by … computer(s) [The Examiner asserts that it is optimal to use computers to implement the algorithm, PIFu. Further, it is also optimal to store algorithms on permanent forms of storage, such as hard drives, diskettes, CD-ROMs, etc., which are analogous to non-transitory computer-readable storage media encoded with instructions.], cause the … computer(s) to perform operations comprising … ([The remaining limitations are repeated verbatim from those recited in independent claim 1.]).
Regarding claim 15, SAITO-2020 discloses a system comprising … computer(s) and … storage device(s) storing instructions (SAITO-2020; p. 82, left col.; “We base our method on the recently introduced Pixel-Aligned Implicit Function (PIFu) representation.” [The Examiner asserts that PIFu is an algorithmic method, which is analogous to ‘instructions’.]) that are operable, when executed by the … computer(s) [The Examiner asserts that it is optimal to use computers to implement the algorithm, PIFu.], to cause the … computer(s) to perform operations comprising: … ([The remaining limitations are repeated verbatim from those recited in independent claim 1.]).
Regarding claim 3 and claim 17, SAITO-2020 discloses the method of claim 1 and the system of claim 15, wherein the observation set comprises respective observations of each of multiple surfaces of a body of the object (SAITO-2020; FIG. 2; p. 83, left col.; “To … improve the quality and fidelity of the reconstruction, we first predict normal maps for the front/back sides [of the person] in image space, and feed these to the network as additional input.”).
Regarding claim 4 and claim 18, SAITO-2020 discloses the method of claim 3 and the system of claim 17, wherein the multiple surfaces comprise a front surface and a back surface of the object relative to a fixed camera (SAITO-2020; p. 83, right col., 1st paragraph; “… the function ‘f’ first extracts an image feature embedding from the projected 2D location at π(X)=x ∈ R2, which we denote by Φ(x, I). Orthogonal projection is used for π, and thus x = π(X) = (Xx, Xy). Then, it estimates the occupancy of the query 3D point X …” [The Examiner notes that the instant specification discloses “π represents a fixed camera, e.g., the camera that captured the image I” {p. 11, line 13}.]).
Regarding claim 5 and claim 19, SAITO-2020 discloses the method of claim 3 and the system of claim 17, wherein the observation for each of the surfaces of the object comprises
a surface normal image corresponding to the surface (SAITO-2020; FIG. 2; p. 83, left col.; “To further improve the quality and fidelity of the reconstruction, we first predict normal maps for the front and back sides in image space, and feed these to the network as additional input.”)
.
Regarding claim 6 and claim 20, SAITO-2020 discloses the method of claim 1 and the system of claim 15, wherein generating a [3-D] model of the object from the final observation set comprises:
generating, from the final observation set, features of the [3-D] model of the object (SAITO-2020; FIG. 2; § 3.2; pp. 83-84; “We present a multi-level approach towards higher fidelity 3-D human digitization that takes 1024×1024 resolution images as input. Our method is composed of two levels of PIFu modules: (1) a coarse level [like] PIFu, focusing on integrating global geometric information by taking the downsampled 512 × 512 image as input, and producing backbone image features of 128 × 128 resolution …” [Coarse PIFu features]);
determining a neural implicit surface from the features (SAITO-2020; § 1; p. 82, left col.; “… for a complete reconstruction, the system needs to recover the backside, which is unobserved in any single image. As with low resolution input, missing information that is not predictable from observable measurements will result in overly smooth and blurred estimates. We overcome this problem by leveraging image-to-image translation networks to produce backside normals … Conditioning our multi-level pixel-aligned shape inference with the inferred back-side surface normal [neural implicit surface] removes ambiguity and significantly improves the perceptual quality of our reconstructions with a more consistent level of detail between the visible and occluded parts. The main contribution in this work consists of an end-to-end trainable coarse-to-fine framework for implicit surface learning for high-resolution 3-D clothed human reconstruction …”); and
rendering a set of points using the neural implicit surface to generate an estimate of the [3-D] representation of the body of the object (SAITO-2020; FIG. 2, rightmost sub-image ‘Reconstruction’; § 3.1; p. 83, left col.; “… the foundation of PIFu …, which constitutes the coarse level of our method (upper half in Fig. 2). The goal of 3-D human digitization [is] achieved by estimating the occupancy of a dense 3-D volume, which determines whether a point in 3-D space is inside the human body or not. In contrast to previous approaches, where the target 3-D space is discretized and algorithms focus on estimating the occupancy of each voxel …, the goal of PIFu is to model a function, f(X), which predicts the binary occupancy value for any given 3-D position in continuous camera space X =(Xx, Xy, Xz) ∈ R3 … [See Equation 1, which defines the ‘binary occupancy value’ based on the ‘continuous camera space’ and the RGB image I.]” § 3.1; p. 83, right col.; “Unlike voxel-based methods, PIFu does not produce a discretized volume as output, so training [is] performed by sampling 3-D points and computing the occupancy loss at the sampled locations, without generating 3-D meshes. During inference, 3-D space is uniformly sampled to infer the occupancy, and the final iso-surface is extracted with a threshold … using marching cubes …”).
Regarding claim 11, SAITO-2020 discloses the method of claim 1, wherein, at each sampling iteration, updating the observation set using the features comprises:
processing the features using a generator neural network to generate an updated observation set (SAITO-2020; p. 83, left col.; “PIFu models the function f via a neural network architecture that is trained in an end-to-end manner.” p. 83, right col.; “… a … (CNN) architecture is used for the 2-D feature embedding function Φ and a Multilayer Perceptron (MLP) for the function g.”).
Regarding claim 12, SAITO-2020 discloses the method of claim 1, wherein generating, from the initial updated observation set, features of the [3-D] model of the object comprises:
processing the initial updated observation set and the conditioning image using a feature extractor neural network to generate the features (SAITO-2020; FIG. 2; [The Examiner notes that the feature extractor of the coarse PIFu processes the initial updated observation set and the conditioning image.]).
Regarding claim 13, SAITO-2020 discloses the method of claim 1, wherein the features are pixel-aligned features in a space of the conditioning image (SAITO-2020; FIG. 2; § 1, ¶ 4, p. 82, left col.; “… we introduce an end-to-end multi-level framework that infers 3D geometry of clothed humans … in a pixel aligned manner, retaining the details in the original inputs without any post-processing. Our method differs from the coarse-to-fine approaches in that no explicit geometric representation is enforced in the coarse levels. Instead, implicitly encoded geometrical context is propagated to higher levels without making an explicit determination about geometry prematurely. We base our method on the recently introduced Pixel-Aligned Implicit Function (PIFu) representation. The pixel-aligned nature of the representation allows us to seamlessly fuse the learned holistic embedding from coarse reasoning with image features learned from the high-resolution input in a principled manner. Each level incrementally incorporates additional information missing in the coarse levels, with the final determination of geometry made only in the highest level.”; § 3.1 ‘Pixel-Aligned Implicit Function’).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 USC 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 2 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over SAITO-2020 as applied to claims 1 and 15 above, respectively, and further in view of Moghaddam et al. (U.S. PG-PUB 2026/0178919, 'MOGHADDAM').
Regarding claim 2 and claim 16, SAITO-2020 discloses the method of claim 1 and the system of claim 15; however, SAITO-2020 does not explicitly disclose that initializing the observation set comprises:
sampling each value in the observation set from a respective noise distribution, which MOGHADDAM discloses (MOGHADDAM; FIG. 1; ¶ 0063; “GANs can generate images from random noise and do not require detailed information or labels from existing samples to begin the generative process. The standard GAN structure consists of two neural networks: a generator G and a discriminator D … The generator G takes random noise … sampled from a uniform or normal distribution as input and maps the noise variable … to the data space x=G(z). The discriminator ‘D’ distinguishes whether an image is real or fake (i.e., made by the generator). The output D(x) is the probability that the input x is real. If the input is a fake image, D(x) would be zero. Through this process, the discriminator ‘D’ is trained to maximize the probability of assigning the correct label to both real samples and fake samples. Generator G is encouraged simultaneously to fit the true data distribution.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person having ordinary skill in the art to modify the method of claim 1 and the system of claim 15 of SAITO-2020 to include the sampling each value in the observation set from a respective noise distribution of MOGHADDAM. The motivation for this modification is to build and validate a GAN-based generative design model with an offline design evaluation function to generate samples that are not only realistic, but also diverse and desirable. A multimodal Data-driven Design Evaluation (DDE) model is provided to guide the generative process by automatically predicting user sentiments for the generated samples based on large-scale user reviews of previous designs. DDE can be incorporated into the StyleGAN structure, a GAN model, to enable data-driven generative processes that are innovative and user-centered. The results of experiments conducted on a large dataset of images of footwear products demonstrate the effectiveness of the proposed DDE-GAN in generating high-quality and diverse concepts (MOGHADDAM, ¶ [0005]).
Claims 7 and 9 are rejected under 35 U.S.C. 103 as being unpatentable over SAITO-2020 as applied to claim 6 above, and further in view of Takikawa et al. (U.S. PG-PUB 2022/0172423, 'TAKIKAWA').
Regarding claim 7, SAITO-2020 discloses the method of claim 6; however, SAITO-2020 does not explicitly disclose that rendering the set of points comprises:
rendering the set of points using sphere tracing, which TAKIKAWA discloses (TAKIKAWA; ¶ 23-24; “Advanced geometric modeling and rendering techniques in computer graphics can utilize 3-D shapes with complex details, arbitrary topology, and quality, usually leveraging polygon meshes. However, it is non-trivial to adapt those representations to learning-based approaches since they lack differentiability and thus cannot easily be used in computer vision applications such as learned image-based 3D reconstruction. … neural approximations of signed distance functions (neural SDFs) provide an attractive choice to scale up computer vision and graphics applications. Neural networks can encode accurate 3D geometry without restrictions on topology or resolution by learning the appropriate SDF, which can define a surface by its zero level-set. A … fixed-size multi-layer perceptron (MLP) can be utilized as the learned distance function. Directly rendering and probing neural SDFs can utilize an approach such as sphere tracing, a root-finding algorithm that can require hundreds of SDF evaluations per pixel to converge. … small neural network(s) [are] utilized to overfit single shapes, but this … comes at the cost of generality and reconstruction quality. Fixed-size neural networks may be utilized, but these may then be unable to express geometry with complexity exceeding the [network capacity]. Accordingly, approaches … can utilize representations for neural SDFs that can adaptively scale to [distinct] levels of detail (LODs) and reconstruct highly detailed geometry. Such an approach can smoothly interpolate between different scales of geometry and can be rendered in real-time with a reasonable memory footprint.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person having ordinary skill in the art to modify the method of claim 6 of SAITO-2020 to include rendering the set of points using sphere tracing of TAKIKAWA. The motivation for this modification is to implement a multi-layer perceptron (MLP) that can be used to make sphere tracing practical without sacrificing quality or generality. Such an approach can build on surface extraction mechanisms that utilize quadrature and spatial data structures storing distance values to finely discretize the Euclidean space such that simple, linear basis functions can reconstruct the geometry (TAKIKAWA; ¶ 0024).
Regarding claim 9, SAITO-2020 discloses the method of claim 6; however, SAITO-2020 does not explicitly disclose that determining the neural implicit surface from the features comprises:
determining the neural implicit surface using a signed distance function neural network that … receive(s) an input derived from the features and an input point, which TAKIKAWA discloses (TAKIKAWA; ¶ 0021; “… neural signed distance functions (SDFs) provide an effective representation for … (3D) shapes, particularly useful for graphics applications. Neural SDFs are functions of position, which return the nearest distance to the surface for any coordinate, such as may be given by ƒ (x, y, z) [input point] = d. Neural SDFs have the benefit of being differentiable and smooth, although they may be … slow to render in … existing systems. Conventional methods typically encode an SDF using a large, fixed-size neural network to approximate complex shapes with implicit surfaces.”) and to generate an output that estimates a signed distance of the input point from the neural implicit surface (TAKIKAWA; ¶ 0023-24; “Advanced geometric modeling and rendering techniques in computer graphics can utilize 3-D shapes with complex details, arbitrary topology, and quality, usually leveraging polygon meshes. However, it is non-trivial to adapt those representations to learning-based approaches since they lack differentiability and thus cannot easily be used in computer vision applications such as learned image-based 3D reconstruction. … neural approximations of signed distance functions (neural SDFs) provide an attractive choice to scale up computer vision and graphics applications. Neural networks can encode accurate 3D geometry without restrictions on topology or resolution by learning the appropriate SDF, which can define a surface by its zero level-set. A … fixed-size multi-layer perceptron (MLP) can be utilized as the learned distance function. Directly rendering and probing neural SDFs can utilize an approach such as sphere tracing [generate an output that estimates a signed distance of the input point from the neural implicit surface], a root-finding algorithm that … require hundreds of SDF evaluations per pixel to converge. … small neural network(s) may be utilized to overfit single shapes, but this often comes at the cost of generality and reconstruction quality. Fixed-size neural networks may be utilized, but these may then be unable to express geometry with complexity exceeding the capacity of the network. Accordingly, approaches … can utilize representations for neural SDFs that can adaptively scale to [distinct] levels of detail (LODs) and reconstruct highly detailed geometry. Such an approach can smoothly interpolate between different scales of geometry and can be rendered in real-time with a reasonable memory footprint.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person having ordinary skill in the art to modify the method of claim 6 of SAITO-2020 to include the determining the neural implicit surface using a signed distance function neural network that … receive(s) an input derived from the features and an input point and to generate an output that estimates a signed distance of the input point from the neural implicit surface of TAKIKAWA. The motivation for this modification is to implement signed distance functions (SDFs), which are used to describe object geometry in real-time rendering, usually in a ray-marching context. Further, SDFs can be implemented as a loss function to minimize the error in interpenetration of pixels while rendering multiple objects. In particular, for any pixel that does not belong to an object, if it lies outside the object in rendition, no penalty is imposed; if it does, a positive value proportional to its distance inside the object is imposed.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over SAITO-2020 as applied to claim 6 above, and further in view of Museth et al. (U.S. PG-PUB 2004/0170302, 'MUSETH').
Regarding claim 8, SAITO-2020 discloses the method of claim 6; however, SAITO-2020 does not explicitly disclose that rendering the set of points comprises:
extracting a mesh from the set of points; and
rasterizing the extracted mesh, both of which MUSETH discloses (MUSETH; FIG. 1, ‘Volume Rendering Mesh Extraction 32’; ¶ 0215; “… level set surfaces [are] rendered … indirectly by a simple two-step procedure (a polygonal mesh is extracted from the volume dataset and rasterized on graphics hardware). [This] has been found to perform and scale better with the size of the test volumes. Implementing a few straightforward mesh extraction procedures make the overhead of the indirect rendering approach insignificant. Conventional graphics hardware is then capable of providing interactive framerates for [all] models …”).
Before the effective filing date of the claimed invention, it would have been obvious to a person having ordinary skill in the art to modify the method of claim 6 of SAITO-2020 to include the extracting a mesh from the set of points and the rasterizing the extracted mesh of MUSETH. The motivation for this modification is to implement a method and/or a system for interactive editing of geometric models based on level set operators (MUSETH; ¶ [0004]).
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over SAITO-2020*TAKIKAWA as applied to claim 9 above, and further in view of D'Innocente et al. (U.S. Patent 11,809,520; 'DINNOCENTE').
PNG
media_image3.png
590
727
media_image3.png
Greyscale
Regarding claim 10, SAITO-2020*TAKIKAWA disclose the method of claim 9; however, SAITO-2020*TAKIKAWA do not explicitly disclose that the input to the signed distance function neural network comprises a feature vector for the input point that is generated by projecting the input point onto an image plane to generate a pixel location, which DINNOCENTE discloses (SAITO-2020; p. 83, right col.; [See screen capture above; the Examiner asserts that a feature vector for the input point is analogous to the ‘image feature embedding’.]); however, SAITO-2020*TAKIKAWA do not explicitly disclose bilinearly interpolating the features at the pixel location, which DINNOCENTE discloses (DINNOCENTE; FIG. 1; Col. 5, Lines 24-36; “… the point of interest lies in a space of the feature maps 112 that is bounded by four neighboring grid points on the C×C feature maps 112 (e.g., four coordinates in the feature maps 112 defining a rectangle that corresponds to a region surrounding the first point of interest 108 in the pixel space). Although often described herein as a single point of interest, a location-of-interest of any size may instead be used. … a group of pixels may be selected (e.g., by drawing a bounding box). … the feature vectors at these four nearest grid points to the location of interest may be bilinearly interpolated (e.g., using bilinear interpolation 116) to generate an approximation of the feature representation at the point of interest 108.”).
Before the effective filing date of the claimed invention, it would have been obvious to a person having ordinary skill in the art to modify the method of claim 6 of SAITO-2020*TAKIKAWA to include the bilinearly interpolating the features at the pixel location of DINNOCENTE. The motivation for this modification is to implement techniques that learn localized representations for fashion retrieval based on user-specified local interest points of prominent visual features present in the images (DINNOCENTE; Col. 1, Lines 59-62).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JONATHAN M COFINO whose telephone number is (303) 297-4268. The examiner can normally be reached Monday-Friday 10A-4P MT.
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, Kent Chang can be reached at 571-272-7667. 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.
/JONATHAN M COFINO/Examiner, Art Unit 2614
/KENT W CHANG/Supervisory Patent Examiner, Art Unit 2614
Prosecution Insights