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 .
DETAILED ACTION
Priority
No foreign or domestic priority is claimed. The effective filing date of U.S. Application No. 18/623,979 is 04/01/2024.
Status of Claims
Claims 1–3, 5-21 are pending in the application.
Claim 4 is cancelled. Claims 1-3, 6-11, 21 are rejected.
Claim 5 is objected to.
Claims 12-20 are allowed.
Allowable Subject Matter
Claim 5 is objected to as being dependent upon a rejected base claim(s), but would be allowable if rewritten in independent form including all of the limitations of the base claim(s) and any intervening claim(s).
Overview of Grounds of Rejection
Ground of Rejection
Claim(s)
Statute(s)
Reference(s)
1
1, 3, 8-11, 21
§ 103
Liu et al. (NPL) in view of Hou et al. (NPL)
2
2, 7
§ 103
Liu et al. (NPL) in view of Hou et al. (NPL), and further in view of Ding et al. (NPL)
3
6
§ 103
Liu et al. (NPL) in view of Hou et al. (NPL), and further in view of Yoo et al. (NPL)
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.
(Please see the cited paragraphs, sections, pages, or surrounding text in the references for the paraphrased content.)
Ground of Rejection 1
Claims 1, 3, 8, 9, 10, 11, 21 are rejected under 35 U.S.C. § 103 as being unpatentable over Liu et al. (NPL) in view of Hou et al. (NPL).
Claim 1 is rejected under 35 U.S.C. § 103 as being unpatentable over Liu et al. (NPL) in view of Hou et al. (NPL).
As per Claim 1, Liu et al. (NPL) teaches the following portion of Claim 1, which recites:
“A method comprising: establishing, for a neural texturing model, a value-per-color (VPC) palette”
Liu et al. (NPL) teach a neural texturing model that uses a learned color palette with per-location selection values:
“Given an input image, our shape and color generators generate a triangle mesh M and its corresponding colors C, which are then fed into the soft rasterizer.” — Liu et al. (NPL), Sec. 4.1, p. 5.
“We propose a novel approach to colorize mesh using a color palette… a sampling network that samples the representative colors for building the palette, and a selection network that combines colors from the palette for texturing the sampling points. The color prediction is obtained by multiplying the color selections with the learned color palette.” — Liu et al. (NPL), Sec. 4.1.1, p. 5.
These passages show a neural texturing model (mesh plus colors output by neural generators) and a palette with selection values per palette color per sampling point, matching “establishing… a value-per-color (VPC) palette” for the neural texturing model.
Liu et al. further teach the following newly added portion of Claim 1, which recites:
“providing, as input to the neural texturing model, a coordinate representing a location on a computerized object”
Liu et al. teach that the neural rendering/texturing pipeline operates on mesh and image-space locations. Liu et al. state that “our shape and color generators generate a triangle mesh M and its corresponding colors C, which are then fed into the soft rasterizer” and further describe a color generator in which a selection network “combines colors from the palette for texturing the sampling points,” with “color prediction … obtained by multiplying the color selections with the learned color palette.” — Liu et al. (NPL), Sec. 4.1 and Sec. 4.1.1, p. 5. Liu et al. also state that, “[f]ollowing the standard rendering pipeline,” the system obtains “image-space coordinate U” by transforming input geometry M based on camera P. — Liu et al. (NPL), Sec. 3.1, p. 3. Further, Liu et al. represent image-space/pixel locations using coordinate form pᵢ = [x; y; 1]ᵀ. — Liu et al. (NPL), App. A2.1.
Thus, Liu’s sampling points / image-space pixel locations correspond to coordinates representing locations on a computerized object, and the neural color generator processes such location information to generate color selections for those locations.
Liu alone does not explicitly teach all the limitation(s) of the claim. However, when combined with Hou et al. (NPL), they collectively teach all of the limitation(s).
Hou et al. teach the following portion of Claim 1, which recites:
“…with a highest value at a respective location of the neural texturing model determining which color is selected for the respective location”
Hou et al. (NPL) provide the “highest value selects color” behavior for a palette:
“Two depth-wise… convolution layers create a softmax probability map of each pixel taking one specific color. This results in a C-channel probability map m(x)… the 1-channel color index map M(x) is computed as the arg max over the C-channel probability map m(x), M(x) = arg max_c m(x).” — Hou et al. (NPL), Sec. 4.2, p. 3.
Here, each pixel/location has a vector of per-color values m(x), and the arg-max operation selects the palette index whose value is highest, so that “a highest value… determining which color is selected” is directly implemented by M(x) = arg max_c m(x).
Hou et al. teach the following portion of Claim 1, which recites:
“selecting, based on output from the neural texturing model processing the coordinate representing the location on the computerized object as input and from the VPC palette, a respective color for the location on the computerized object.”
Hou et al. (NPL) shows selection from a palette per location:
“First, the convolutional layers output a C-channel probability map m(x) for C colors. Next, a 1-channel color index map M(x) is created via the arg max function. Then, the color palette T(x) is computed as average of all pixels that are of the same color index. At last, the color quantized image x is created via a table look-up session.” — Hou et al. (NPL), Fig. 4 caption, p. 4.
“At last, the quantized image x is created as x = Σ_c [T(x)]_c · I(M(x) = c).” — Hou et al. (NPL), Eq. (5), p. 4.
The index map M(x) identifies, for each pixel/location, a palette index, and the expression x = Σ_c [T(x)]_c · I(M(x) = c) shows that for locations where M(x) = c, the corresponding palette color [T(x)]_c is selected. Thus, the output per-location values from the neural model, together with the VPC/color palette, are used to select a respective color for the corresponding location on the computerized object.
Before the effective filing date of the claimed invention, a person of ordinary skill in the art working on neural rendering and color quantization would have recognized that Liu et al. (NPL) provide a differentiable neural texturing pipeline with a learned color palette and per-location selection values used for coloring a mesh within Soft Rasterizer. Hou et al. (NPL) teach a complementary design in which a network predicts a C-channel probability map m(x) over palette colors, then at test time computes a color index map M(x) = arg max_c m(x) and uses a palette T(x) with table lookup to produce a quantized image, thereby storing compact indices instead of full color vectors. In light of these teachings, a POSITA would have been motivated to keep Liu et al.’s soft, differentiable VPC representation during training, yet adopt Hou et al.’s argmax-based index selection and palette lookup at inference or final rendering so that the highest per-color value at each location selects a single palette entry, yielding sharper, discrete textures and reduced memory usage through indexed palette storage, which represents a predictable and routine integration of known techniques. Applying Hou et al.’s per-location probability-map and palette-lookup mechanism to Liu et al.’s neural mesh color generator would have predictably resulted in a system where coordinates or sampling locations on a computerized object are processed by the neural texturing model to produce per-color values, and the resulting model output, together with the VPC palette, is used to select a color for that location.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
As per Claim 3, Liu alone does not explicitly teach all the limitation(s) of the claim. However, when combined with Hou et al. (NPL), they collectively teach all of the limitation(s) of Claim 3 that recites:
“The method of claim 1, wherein the highest value is a highest value that is differentiable.”
For Claim 1, the “highest value” is implemented as M(x) = arg max_c m(x) in Hou’s ColorCNN. Hou then provides a differentiable replacement of this argmax during training:
“To start with, we remove the arg max 1-channel color index map M(x) in Eq. 3. Instead, we use the C-channel softmax probability map m(x). Next, we change the color palette design… For each quantized color… we set its RGB color value [t(x)]_c as the weighted average over all pixels… At last, instead of table look-up, the quantized image x̃ is computed as the weighted average of all colors in the color palette.” — Hou et al. (NPL), Sec. 4.3.1 “Differentiable Approximation,” Fig. 5, p. 4.
Here the C-channel softmax probability map m(x) is a differentiable function of the network outputs, and its entries act as per-color “values” that control each color’s contribution.
Replacing the discrete argmax “highest value” with this softmax-based, differentiable value vector makes the effective “highest value” operation differentiable in the sense claimed.
Before the effective filing date of the claimed invention, a person of ordinary skill in the art would have understood from Hou et al. (NPL) that a non-differentiable argmax over palette values can be replaced during training by a C-channel softmax probability map m(x) to obtain a differentiable forward pass. In view of these teachings, a POSITA would have found it routine to apply a softmax-based relaxation to the “highest value” selection in the combined Liu + Hou neural texturing model, thereby making the claimed highest value differentiable with predictable benefits for gradient flow and training stability.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
Processor system Claim 8 is rejected under 35 U.S.C. § 103 as being unpatentable over Liu et al. (NPL) in view of Hou et al. (NPL).
Claim 8 recites:
“A processor system configured to: for each of at least some respective locations on a computerized object, input respective coordinates of one or more respective locations on a computerized object to at least one machine learning (ML) model; receive, as output from the ML model corresponding to the respective coordinates input to the ML model, respective colors for the respective locations; and cause the computerized object to be rendered on at least one display using the respective colors received from the ML model for each respective location of the computerized object.”
As per Claim 8, Liu et al. (NPL) teach a processor-implemented neural rendering/texturing system in which coordinates/locations on a computerized object are processed to generate colors for rendering:
“Given an input image, our shape and color generators generate a triangle mesh M and its corresponding colors C, which are then fed into the soft rasterizer.” — Liu et al. (NPL), Sec. 4.1, p. 5.
Liu et al. further teach that the rendering pipeline uses coordinate-based locations, including “image-space coordinate U” obtained by transforming geometry M based on camera P, and image-space/pixel coordinates such as pᵢ = [x; y; 1]ᵀ. — Liu et al. (NPL), Sec. 3.1, p. 3; Appendix / Supplemental Materials, Sec. A2.1.
Liu et al. also teach a color generator in which a selection network “combines colors from the palette for texturing the sampling points,” and “color prediction is obtained by multiplying the color selections with the learned color palette.” — Liu et al. (NPL), Sec. 4.1.1, p. 5.
Hou et al. (NPL) teach receiving respective colors for locations by generating a per-location C-channel probability map m(x), computing a color index map M(x), and creating the final image by palette lookup: “the color quantized image x is created via a table look-up session.” — Hou et al. (NPL), Sec. 4.2, Fig. 4 caption, p. 4.
Thus, Liu in view of Hou teaches or suggests a processor system configured to input respective coordinates/locations on a computerized object to an ML/neural texturing model, receive respective colors for those locations, and render the computerized object using the respective colors, as recited in Claim 8.
Before the effective filing date, a POSITA would have been motivated to apply Hou et al.’s per-
location color-output and palette-lookup technique to Liu et al.’s neural rendering/texturing system to obtain predictable per-location colors for rendering a computerized object, thereby arriving at Claim 8.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
As per Claim 9, Liu et al. (NPL) teaches the additional limitation of Claim 9 that recites:
“The processor system of Claim 8, wherein the coordinates comprise Cartesian coordinates.”
Liu et al. (NPL) use standard Cartesian (x, y) image coordinates:
“We represent pᵢ = [x; y; 1]ᵀ … where Uⱼ = [x₁ x₂ x₃; y₁ y₂ y₃; …].” — Liu et al. (NPL), Appendix / Supplemental Materials, Sec. A2.1.
This shows that the processor system in Liu’s rendering pipeline uses Cartesian coordinates (x, y) for locations, satisfying Claim 9’s requirement that “the coordinates comprise Cartesian coordinates.”
The rationale and motivation to combine the references as set forth for Claim 8 are incorporated herein by reference for the present claim.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
As per Claim 10, Liu et al. (NPL), as applied in the rejection of Claim 8, further teaches or suggests the additional limitation of Claim 10 that recites:
“The processor system of Claim 8, wherein the computerized object comprises a face mask.”
Liu et al. present Soft Rasterizer as a general-purpose differentiable renderer for arbitrary meshes, with specific motivation for “3D face reconstruction” and “non-rigid objects, e.g. human face.” — Liu et al. (NPL), Sec. 2.
A POSITA, seeing that Liu’s system is expressly applicable to face-related 3D reasoning and operates on generic mesh objects, would have regarded using a mesh representing a face mask as a routine, predictable choice of computerized object shaped for the human face. Thus, specifying that “the computerized object comprises a face mask” would have been an obvious variation.
The rationale and motivation to combine the references as set forth for Claim 8 are incorporated herein by reference for the present claim.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
As per Claim 11, Liu alone does not explicitly teach all the limitation(s) of the claim. However, when combined with Hou et al. (NPL), they collectively teach all of the limitation(s) of Claim 11 that recites:
“The processor system of claim 8, wherein the at least one ML model comprises a multilayer perceptron.”
In the SoftRas system, Liu et al. (NPL) describe several neural components made of stacked fully connected layers:
“The shape generator consists of three fully connected layers and outputs a per-vertex displacement vector…” and “The color generator contains two fully connected streams: one for sampling the input image to build the color palette and the other one for selecting colors from the color palette to texture the sampling points.” — Liu et al. (NPL), App. C, Fig. C3, p. 14.
A POSITA would recognize these stacked fully connected layers as a multilayer perceptron (MLP), so the ML model in Liu’s processor system comprises a multilayer perceptron.
The rationale and motivation to combine the references as set forth for claim 1 are incorporated herein by reference for the present claim.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
As per Claim 21, Liu et al. (NPL) in view of Hou et al. (NPL) teaches the limitation of Claim 21 that recites:
“The processor system of Claim 8, wherein the ML model is trained to output one or more colors based on at least one coordinate of a computerized object as input.”
Liu et al. (NPL) teach neural shape and color generators that output colors for a computerized mesh object:
“Given an input image, our shape and color generators generate a triangle mesh M and its corresponding colors C, which are then fed into the soft rasterizer.” — Liu et al. (NPL), Sec. 4.1, p. 5.
Liu et al. further teach that the color generator textures coordinate-based sampling points: a selection network “combines colors from the palette for texturing the sampling points,” and “color prediction is obtained by multiplying the color selections with the learned color palette.” — Liu et al. (NPL), Sec. 4.1.1, p. 5.
Liu et al. also teach that the model is trained, because “the reconstruction networks are supervised by three losses,” and “the final loss is a weighted sum of the three losses: L = Ls + λLc + μLg.” — Liu et al. (NPL), Sec. 4.1.1, p. 5.
Hou et al. (NPL) teach generating per-location color outputs from coordinate/pixel locations by producing a C-channel probability map m(x) and using palette lookup to form the final color image. — Hou et al. (NPL), Sec. 4.2, Fig. 4 caption, p. 4.
Thus, Liu in view of Hou teaches or suggests an ML model trained to output one or more colors based on at least one coordinate/location of a computerized object as input. The rationale and motivation to combine the references as set forth for Claim 8 are incorporated herein by reference for the present claim.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
Ground of Rejection 2
Claims 2, 7 are rejected under 35 U.S.C. § 103 as being unpatentable over Liu et al. (NPL) in view of Hou et al. (NPL), and further in view of Ding et al. (NPL).
As per Claim 2, Liu alone does not explicitly teach all the limitation(s) of the claim. However, when combined with Ding et al. (NPL), they collectively teach all of the limitation(s) of Claim 2 that recites:
“The method of Claim 1, comprising: generating the VPC value-per-color palette at least in part by making the VPC value-per-color palette to be a trainable parameter such that the VPC value-per-color palette is learned.”
Limitation: “generating the VPC value-per-color palette … such that the VPC value-per-color palette is learned”
Liu et al. (NPL) provide a learned palette inside a neural color generator:
“We propose a novel approach to colorize mesh using a color palette… a sampling network that samples the representative colors for building the palette, and a selection network that combines colors from the palette for texturing the sampling points. The color prediction is obtained by multiplying the color selections with the learned color palette.” — Liu et al. (NPL), Sec. 4.1.1, p. 5.
This shows that the palette is learned as part of the neural color generator.
Limitation: “making the VPC value-per-color palette to be a trainable parameter”
Ding et al. (NPL) provide the missing detail that the palette itself is a trainable parameter set, rather than only a derived output:
“The image tokenizer is a discrete Auto-Encoder… each d-dimensional vector is quantized to a nearby embedding in a learnable codebook {v₀, …, v_{|V|-1}}, ∀v_k ∈ ℝᵈ.” — Ding et al. (NPL), Sec. 2.2, p. 4.
When describing training strategies, Ding et al. state: “The nearest-neighbor mapping, moving average, where each embedding in the codebook is updated periodically during training as the mean of the vectors recently mapped to it [46].” — Ding et al. (NPL), Sec. 2.2, p. 4.
These passages show that the codebook entries themselves are trainable parameters that are updated during training, not just outputs of another network. A POSITA would understand this learnable codebook as a concrete example of a VPC value-per-color palette implemented as trainable parameters.
In combination with Liu’s palette-based neural texturing, Ding’s learnable codebook corresponds to “making the VPC value-per-color palette to be a trainable parameter”: the palette entries, analogous to per-color values/embeddings, are the learned parameters updated throughout training.
Before the effective filing date of the claimed invention, a person of ordinary skill in the art would have recognized from Liu et al. (NPL) that a neural texturing model can use a learned color palette produced within a neural color generator. Ding et al. (NPL) further teach that an image representation can employ a learnable codebook {v₀, …, v_{|V|-1}}, where each embedding in the codebook is updated during training, making the codebook entries themselves trainable parameters. In view of these teachings, a POSITA would have been motivated to implement the VPC value-per-color palette in Liu’s neural texturing model as a learnable codebook in the style of Ding et al., so that the palette entries are trainable parameters updated during training, thereby generating a palette that is both learned and parameterized with predictable benefits for optimization control and representational flexibility.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
As per Claim 7, Liu alone does not explicitly teach all the limitation(s) of the claim. However, when combined with Ding et al. (NPL), they collectively teach all of the limitation(s) of Claim 7 that recites:
“The method of Claim 1, wherein the VPC value-per-color palette is learned and the method comprises: locking the VPC palette after learning so that value-to-color correspondences do not change.”
Limitation: “the VPC value-per-color palette is learned and the method comprises: locking the VPC palette after learning so that value-to-color correspondences do not change.”
Ding et al. (NPL) describe an image tokenizer with a learnable codebook:
“The image tokenizer is a discrete Auto-Encoder… each d-dimensional vector is quantized to a nearby embedding in a learnable codebook {v₀, …, v_{|V|-1}}, ∀v_k ∈ ℝᵈ.” — Ding et al. (NPL), Sec. 2.2 “Tokenization”, p. 4.
The codebook entries themselves are trained during Stage 1:
“The learning process is then divided into two stages: (1) The encoder φ and decoder ψ learn to minimize the reconstruction loss. (2) A single GPT optimizes the two negative log-likelihood (NLL) losses… As a result, the first stage degenerates into a pure discrete Auto-Encoder, serving as an image tokenizer to transform an image to a sequence of tokens; the GPT in the second stage undertakes most of the modeling task.” — Ding et al. (NPL), Sec. 2.1, p. 3.
So the palette/codebook is learned in Stage 1, satisfying the “VPC value-per-color palette is learned” portion.
Limitation: “locking the VPC palette after learning so that value-to-color correspondences do not change”
The two-stage setup in Ding et al. implies that once Stage 1 has produced a discrete tokenizer, it is then used as a fixed vocabulary for Stage 2:
Stage 1: “pure discrete Auto-Encoder, serving as an image tokenizer to transform an image to a sequence of tokens.”
Stage 2: “the GPT in the second stage undertakes most of the modeling task.” — Ding et al. (NPL), Sec. 2.1, p. 3.
A POSITA would understand that for GPT training to be stable, the tokenizer’s codebook must remain fixed after Stage 1; otherwise, token indices would change meaning during Stage 2. In VPC terms, this corresponds to locking the learned palette so that the mapping from each palette index/value to its color (embedding) no longer changes, i.e., “locking the VPC palette after learning so that value-to-color correspondences do not change.”
Before the effective filing date, a POSITA would have recognized from Ding et al. (NPL) that it is standard in VQ-VAE+GPT architectures to first learn a codebook (palette) in Stage 1 and then reuse that same codebook as a fixed tokenizer in Stage 2, so the index-to-embedding correspondences stay stable while the downstream Transformer is trained. Applying this common “learn-then-lock” pattern to the learned VPC value-per-color palette in the Liu + Hou neural texturing model would be a routine engineering choice to ensure stable, consistent palette indices during later training or inference.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
Ground of Rejection 3
Claim 6 is rejected under 35 U.S.C. § 103 as being unpatentable over Liu et al. (NPL) in view of Hou et al. (NPL), and further in view of Yoo et al. (NPL).
As per Claim 6, Liu alone does not explicitly teach all the limitation(s) of the claim. However, when combined with Yoo et al. (NPL), they collectively teach all of the limitation(s) of Claim 6 that recites:
“The method of Claim 1, wherein the VPC value-per-color palette is not learned.”
Yoo et al. (NPL) describe exactly such non-learned palette extraction methods in the GIF context:
“Finding an optimal color palette, which is equivalent to clustering, is an NP-hard problem. A commonly used algorithm for palette extraction is the median-cut algorithm [15], due to its low cost and relatively high quality. Better clustering algorithms such as k-means produce palettes of higher quality, but are much slower… Nearly all classical palette extraction methods involve an iterative procedure over all the image pixels.” — Yoo et al. (NPL), Sec. 1 “Introduction”.
These median-cut and k-means palettes are computed algorithmically from image pixels and are not learned by back-propagation or treated as neural parameters. They are fixed for given input/images and are not optimized jointly with the rest of a network, matching “VPC value-per-color palette is not learned” in the neural-network sense.
Before the effective filing date, a person of ordinary skill in the art would have seen in Hou et al. (NPL) that learned palettes, such as ColorCNN, and non-learned palettes, such as MedianCut and OCTree, were treated as comparable alternatives, since ColorCNN is directly evaluated against “MedianCut + Dither,” “MedianCut,” and “OCTree” baseline quantizers. Yoo et al. (NPL) likewise describe median-cut and k-means as standard, non-learned palette extraction methods. In view of these teachings, choosing a non-learned VPC value-per-color palette instead of a learned one in the Liu + Hou pipeline would have been a routine design choice with predictable trade-offs.
PNG
media_image1.png
13
460
media_image1.png
Greyscale
Response to Arguments
Applicant’s arguments have been fully considered. The arguments are persuasive in part, and certain claims have been allowed or objected to as allowable if rewritten in independent form. However, the arguments are not persuasive as to all pending claims, and the rejections of certain claims are maintained for the reasons set forth above.
Examiner’s Suggested Amendments
If applicant elects to file a formal after-final response, the following amendment is suggested as a possible means to place the application in better condition for allowance and to potentially overcome the rejection under 35 U.S.C. § 103.
It is suggested that claim 8 be amended to include additional narrowing limitations, such as one or more of the limitations recited in claims 9–11 and 21, for further consideration.
Conclusion
The prior art made of record and relied upon in this action is as follows:
Patent Literature:
(none)
Non-Patent Literature (NPL):
Ding et al. — “CogView: Mastering Text-to-Image Generation via Transformers”, 2021. Available at: [https://proceedings.neurips.cc/paper/2021/file/a4d92e2cd541fca87e4620aba658316d-Paper.pdf]
Liu et al. — “Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning”, 2019. Available at: [https://arxiv.org/pdf/1904.01786]
Hou et al. — “Learning to Structure an Image with Few Colors”, 2020. Available at: [https://arxiv.org/pdf/2003.07848]
Yoo et al. — “GIFnets: Differentiable GIF Encoding Framework”, 2020. Available at: [https://openaccess.thecvf.com/content_CVPR_2020/papers/Yoo_GIFnets_Differentiable_GIF_Encoding_Framework_CVPR_2020_paper.pdf]
Note: A PDF copy of each NPL reference is attached with this Office Action. URLs are included for applicant convenience. If a link becomes unavailable in the future, the citation information may be used to locate the reference or access archived versions via the Wayback Machine.
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure and is listed as follows:
Patent Literature:
(none)
Non-Patent Literature (NPL):
(none)
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 ADEEL BASHIR whose telephone number is (571) 270-0440. The examiner can normally be reached Monday-Thursday.
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 Hajnik can be reached on (571) 276-7642. 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.
/ADEEL BASHIR/
Examiner, Art Unit 2616
/DANIEL F HAJNIK/Supervisory Patent Examiner, Art Unit 2616
Prosecution Insights