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 .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim 15 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because Claim 15 is directed to a program per se. Such claimed computer programs do not define any structural and functional interrelationships between the computer program and other claimed elements of a computer which permit the computer program’s functionality to be realized. In contrast, a claimed computer-readable medium encoded with a computer program is a computer element which defines structural and functional interrelationships between the computer program and the rest of the computer which permit the computer program’s functionality to be realized, and is thus statutory. See Lowry, 32 F.3d at 1583-84, 32 USPQ2d at 1035. Since a computer program is merely a set of instructions capable of being executed by a computer, the computer program, per se, is nonstatutory.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are: “a parameter creation section that” and “a dataset creation section that” in claim 14.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 102
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 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 1-15 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Tagra et al. (US Pub. 2022/0121839), hereinafter Tagra.
Regarding claim 1, Tagra discloses a data creation method performed by a data creation device, the data creation method comprising: a step of, by partially or wholly changing a creation parameter which is obtained by conversion of a source image of a freely-selected face to a numerical value, creating a larger number of input creation parameters than a predetermined number from the predetermined number of the creation parameters of the source images (Fig. 4A; Fig. 5; Paragraph [0055]: Process 400 includes, at preliminary stage 410, resizing an input image 412. Resizing may include rescaling input image 412 into different resolutions to form an image pyramid 414, which is a set of images corresponding to the input image 412 but having different resolutions. Image pyramid 414 is input into MTCNN such that each image of image pyramid 414 is run through the three network stages of MTCNN. A first stage 420 comprises a proposal network (P-Net), which is a shallow convolutional neural network (CNN) that identifies candidates for the most likely facial regions within the input image 412. Candidates may be calibrated using estimated bounding box regression, and certain overlapped candidates may be merged using non-maximum suppression (NMS). Candidates identified in the first stage 420 are input into a second stage 430, which comprises a refine network (R-Net). R-Net is a CNN that filters out the false candidates. The second stage 430 may also include NMS and bounding box regression to calibrate the determined bounding boxes. A third stage 440 comprises an output network (O-Net), which is a CNN that detects particular details within the facial region candidates remaining after the second stage 430. In exemplary aspects, O-Net outputs a final bounding box 442 for a face and select facial landmarks 444. The facial landmarks output by O-Net are landmarks for the eyes, nose and mouth; Paragraph [0066]: a set of similar facial images are selected from the reference facial images, where the selected facial images are the top n reference facial images that have the most similarity with the base facial image. For embodiments utilizing Euclidean distance, the reference facial images with the smallest distance are selected as most similar to the base facial images. In some embodiments, the number of images selected for the set of similar images (n) may be a value within a range of two and ten. For example, the set of similar facial images may consist of six reference facial images that have been determined to be most similar to the base facial image. FIG. 5 depicts an example set of similar facial images 510 that may be identified for the example base facial image 512); and a step of, by creating a face image data item on a basis of a plurality of the input creation parameters, creating a face image dataset including a plurality of the face image data items (Fig. 5; Fig. 7; Paragraphs [0071]-[0073]: some embodiments of new facial synthesizer 220 further perform triangulation by dividing each similar facial image and the output image space (where the landmark coordinates are the averaged coordinates) into triangular regions. The triangular regions are determined from the detected facial landmarks. In exemplary embodiments, at least one point on each triangular region is the coordinate of a facial landmark. Further, in exemplary embodiments, Delaunay Triangulation is performed, where a convex hull (a boundary that has no concavities) is created and the triangulation that is chosen is the one in which every circumcircle of a triangle is an empty circle. The triangular regions in each similar facial image are warped to the triangular regions in the output image space using affine transformations. Triangulation helps ensure that similar facial regions in the similar facial images fall in the same region of the output space so that they can be combined as described below. FIG. 6C depicts warped similar image 630 that is warped to the output image space using triangulation. FIG. 7 shows the results of this landmark registration and warping process. Specifically, FIG. 7 depicts the example set of similar facial images 510 of FIG. 5 and a set of warped similar facial images 710 that are all warped to the same output image space …After the similar facial images are aligned and warped to an output image space, the similar facial images are combined by averaging pixel intensities of each similar facial image at each pixel within the output image space. In some embodiments, an average pixel intensity for the output image space is computed by summing the corresponding pixels values of all the warped similar facial images and dividing the sum by the number of similar facial images within the set. In some embodiments, a weighted average may be computed. This result of averaging similar facial images is a newly synthesized facial image. Because this new facial image depicts a face that is created from a combination imaged faces, it does not depict an actual person's face. In this way, the new facial image created by new facial image synthesizer 220 does not expose another person's identity. FIG. 8 depicts an example new facial image 800 that is from the example set of similar facial images 510 of FIG. 5…After a new facial image is created from similar facial images, it is combined with the base facial image. As such, new user facial image generator 230 is generally responsible for combining the base facial image and the new facial image. The resulting combination may be referred to herein as a new user facial image; however, it is should be understood that, in some implementations, this combination is based on a base facial image selected from a database rather than a user input facial image. Combining the new facial image with the base facial image helps to preserve some aesthetics, such as complexion, structure, and size, of the base facial image).
Regarding claim 2, Tagra discloses the data creation method according to claim 1, wherein the data creation device creates a plurality of the input creation parameters from one creation parameter by increasing or reducing some or all of a plurality of parameters constituting the creation parameter (Fig. 4A; Paragraph [0056]: the bounding box output from MTCNN is an initial bounding box, and input face detector 210 applies padding around the initial bounding box coordinates to enlarge the bounding box enclosing the face of the input image. In exemplary embodiments, the amount added to the height of the bounding box is the same amount added to the width. For example, padding may include adding five percent of the initial bounding box height and five percent of the initial bounding box width. It is contemplated that other values may be used for padding the bounding box. Utilizing the final bounding box for an input image, a base facial image may be extracted from the input image for further processing by one or more components of identity obfuscation manager 200).
Regarding claim 3, Tagra discloses the data creation method according to claim 1, wherein, by changing a specific parameter constituting the creation parameter, the data creation device creates the input creation parameters (Fig. 2; Paragraphs [0077]-[0079]: Input image updater 240 is generally responsible for updating the input image with the new user facial image generated by new user facial image generator 230. As the input image may often have other aspects of the individual, such as hair and body, that are not part of the generated new user facial image, input image updater 240 operates to help create a seamless blend between the new user facial image and the input image…aspects in which the base facial image is extracted from the input image, input image updater 240 aligns the new user facial image with the face detected in the input image because the face from the input image indicates a natural-appearing alignment of a face with the rest of the input image. This process may be similar to the alignment and morphing process described with respect to the similar facial images. Facial landmarks are extracted and used to form a convex hull over the detected facial landmarks, and triangulation is performed inside the convex hull, using the facial landmark coordinates. In some embodiments, the triangular regions have points corresponding to angles or midpoints along the convex hull, rather than along a facial image boundary box as described in some embodiments of new facial image synthesizer 220. The triangular regions of the convex hull of the new user facial image are warped to the face within the input image…aligning the new user facial image within the input image, embodiments of input image updater 240 blend the new user facial image with the background of the input image. In exemplary embodiments, input image updater 240 uses Poisson blending, which utilizes image gradients such that the gradient of the facial region within the resulting updated input image is the same or almost the same as the gradient of the facial region in the input image. Additionally, the intensity of intensity of the new user facial image may be adjusted so that it is the same as the intensity of the original facial region detected within the input image; Paragraph [0094]: new user facial image generated for a faceless image is added to the input image by input image updater 240. This process includes aligning the new user facial image within the input image and blending the two images together. Because the input image is faceless in this context, input image updater 240 aligns the new user facial image within the input image in a different manner than that previously described for input images having faces. Where the input image is faceless and the new user facial image is created from a selected base facial image, input image updater 240 separates the face from the background in the new user facial image utilizing a foreground and background separation technique).
Regarding claim 4, Tagra discloses the data creation method according to claim 1, wherein, by blending the input creation parameters and the creation parameter of a preset face image prepared in advance, the data creation device creates final input creation parameters (Paragraphs [0077]-[0079]: Input image updater 240 is generally responsible for updating the input image with the new user facial image generated by new user facial image generator 230. As the input image may often have other aspects of the individual, such as hair and body, that are not part of the generated new user facial image, input image updater 240 operates to help create a seamless blend between the new user facial image and the input image…aspects in which the base facial image is extracted from the input image, input image updater 240 aligns the new user facial image with the face detected in the input image because the face from the input image indicates a natural-appearing alignment of a face with the rest of the input image. This process may be similar to the alignment and morphing process described with respect to the similar facial images. Facial landmarks are extracted and used to form a convex hull over the detected facial landmarks, and triangulation is performed inside the convex hull, using the facial landmark coordinates. In some embodiments, the triangular regions have points corresponding to angles or midpoints along the convex hull, rather than along a facial image boundary box as described in some embodiments of new facial image synthesizer 220. The triangular regions of the convex hull of the new user facial image are warped to the face within the input image…aligning the new user facial image within the input image, embodiments of input image updater 240 blend the new user facial image with the background of the input image. In exemplary embodiments, input image updater 240 uses Poisson blending, which utilizes image gradients such that the gradient of the facial region within the resulting updated input image is the same or almost the same as the gradient of the facial region in the input image. Additionally, the intensity of intensity of the new user facial image may be adjusted so that it is the same as the intensity of the original facial region detected within the input image).
Regarding claim 5, Tagra discloses the data creation method according to claim 4, wherein the preset face image includes an imaginary face image (Fig. 2; Paragraph [0053]: Referring to FIG. 2, aspects of an illustrative identity obfuscation manager 200 are shown, in accordance with various embodiments of the present disclosure. At a high level, identity obfuscation manager, which may be implemented within an operating environment as described with respect to identity obfuscation manager 106 in environment 100, facilitates operations for obfuscating facial identity in a user image by synthesizing a new face (embodied as a new facial image) to be combined with the user's input image. Embodiments of identity obfuscation manager 200 includes input face detector 210, new facial image synthesizer 220, new user facial image generator 230, input image updater 240, and data store 250).
Regarding claim 6, Tagra discloses the data creation method according to claim 4, wherein the data creation device performs cleansing on the creation parameters of a plurality of the preset face images, and blends the creation parameters of the preset face images having undergone the cleansing and the input creation parameters (Fig. 2; Paragraphs [0077]-[0079]: Input image updater 240 is generally responsible for updating the input image with the new user facial image generated by new user facial image generator 230. As the input image may often have other aspects of the individual, such as hair and body, that are not part of the generated new user facial image, input image updater 240 operates to help create a seamless blend between the new user facial image and the input image…aspects in which the base facial image is extracted from the input image, input image updater 240 aligns the new user facial image with the face detected in the input image because the face from the input image indicates a natural-appearing alignment of a face with the rest of the input image. This process may be similar to the alignment and morphing process described with respect to the similar facial images. Facial landmarks are extracted and used to form a convex hull over the detected facial landmarks, and triangulation is performed inside the convex hull, using the facial landmark coordinates. In some embodiments, the triangular regions have points corresponding to angles or midpoints along the convex hull, rather than along a facial image boundary box as described in some embodiments of new facial image synthesizer 220. The triangular regions of the convex hull of the new user facial image are warped to the face within the input image…aligning the new user facial image within the input image, embodiments of input image updater 240 blend the new user facial image with the background of the input image. In exemplary embodiments, input image updater 240 uses Poisson blending, which utilizes image gradients such that the gradient of the facial region within the resulting updated input image is the same or almost the same as the gradient of the facial region in the input image. Additionally, the intensity of intensity of the new user facial image may be adjusted so that it is the same as the intensity of the original facial region detected within the input image; Paragraph [0094]: new user facial image generated for a faceless image is added to the input image by input image updater 240. This process includes aligning the new user facial image within the input image and blending the two images together. Because the input image is faceless in this context, input image updater 240 aligns the new user facial image within the input image in a different manner than that previously described for input images having faces. Where the input image is faceless and the new user facial image is created from a selected base facial image, input image updater 240 separates the face from the background in the new user facial image utilizing a foreground and background separation technique).
Regarding claim 7, Tagra discloses the data creation method according to claim 1, wherein the data creation device scrambles the creation parameter of the source image, and creates the input creation parameters on a basis of the scrambled creation parameter (Paragraph [0040]: embodiments of the present invention are directed to facilitating realistic and aesthetically pleasing facial identity obfuscation. At a high level, when an input image is received from or by the direction of a user, a base face for the input image is determined. Face detection technologies are applied to determine if the input image depicts a face, which will be used as the base face. If not, such as when the face is cropped out of the image, is blurred, or is otherwise obscured, the base face is selected from reference facial images. Based on the base face for an input image, a set of similar facial images are selected and used to synthesize a new facial image. The new facial image is combined with the base face, and the input image is updated with the combination of the new facial image and base face, which helps to retain some aesthetics of the input image).
Regarding claim 8, Tagra discloses the data creation method according to claim 7, wherein the data creation device performs the scrambling by adding a random noise to a parameter in a specific layer of a plurality of layers constituting the creation parameter of the source image (Fig. 4B; Paragraphs [0057]-[0060]: FIG. 4B depicts a network architecture for an embodiment of input face detector 210. Specifically, FIG. 4B depicts an example MTCNN architecture 450, which includes a P-Net 452, an R-Net 454, and an O-Net 456. Each network (also referred to as a stage) includes a series of max pooling (designated as “MP” in FIG. 4B) and convolutional (designated as “cony” in FIG. 4B) layers. In one embodiment, the step size in convolution is one, and the step size in pooling two…input face detector 210 detects a face depicted in the input image, the face extracted from the input image may be referred to as a base facial image that is utilized for synthesis of the new facial image as described herein. In some instances, however, input face detector 210 may not detect a face in the input image, such as where the face is cropped out of the input image, is blurred, or is otherwise obscured by an object, such as a block box, in the input image. In some embodiments, if input face detector 210 is unable to detect a face, a message is displayed to the user requesting that a new image depicting a face be input. Additionally, the message may indicate that a new face will be generated for the input image to help protect the user's identity. Where a user does not upload an image with a detectable face (such as where the user initially inputs an image without a face or refuses to input a new image upon receiving a request), some embodiments of the disclosed technology select a base facial image for the input image as described further below with respect to base face selector 270).
Regarding claim 9, Tagra discloses the data creation method according to claim 1, wherein, by performing face ID/attribute labeling and cleansing on the face image dataset, the data creation device creates a final face image dataset (Fig. 10E; Paragraph [0080]: FIGS. 10A-10E depict a series of images as user image is updated with a new user facial image. FIG. 10A depicts a new user facial image 1010 with detected facial landmarks 1012 in a convex hull 1014. FIG. 10B depicts a triangulated new user facial image 1020 that is divided into triangles based on the facial landmarks 1012 and convex hull 1014. FIG. 10C depicts a portion of an aligned updated input image 1030 in which the new user facial image 1010 is aligned with the face region of the user input image. FIG. 10D depicts a portion of blended updated input image 1040 in which the new user facial image and the rest of the updated input image are blended together to create a seamless transition between the original aspects of the input image and the new user facial image that has been created. FIG. 10E depicts the final updated input image 1050 in its entirety).
Regarding claim 10, Tagra discloses the data creation method according to claim 9, wherein, by performing cleansing on the face image dataset, the data creation device creates the final face image dataset that satisfies a predetermined requirement (Paragraph [0097]: identity obfuscation manager 200 include a similarity score determiner 280 that is generally responsible for determining a similarity score for an updated input image. A similarity score is a metric measuring the degree of similarity or the degree of difference between the updated input image and the original input image. In some embodiments, the similarity score is utilized to ensure that some aesthetics of the input image are retained in the updated user input image. In these cases, the similarity score determined for an updated input image is compared to a predetermined minimum threshold similarity score, and if the similarity score for the updated input image does not meet the minimum threshold, a user is provided a notification indicating that identity obfuscation will change the aesthetics of the user's input image. Additionally or alternatively, the similarity score may be utilized to ensure that the updated input image is not too similar to the input image so that the identity is not properly protected. As such, in these instances, the similarity score determined for an updated input image is compared to a predetermined maximum threshold similarity score, and if the similarity score for the updated input image does not meet the maximum threshold, a user is provided a notification indicating that facial identity could not be protected. Where the updated input image does not meet the maximum threshold score, the updated input image may either be presented with the notification or may not be presented at all. In some embodiment, the determined similarity score is compared to both a maximum score and a minimum score to determine whether the updated input image both sufficiently obfuscates the individual's identity and preserves the aesthetics of the original input image. Further, in some embodiments, the similarity score is computed and provided as feedback to other components of identity obfuscation manager 200 to help improve the functioning of identity obfuscation manager 200).
Regarding claim 11, Tagra discloses the data creation method according to claim 10, wherein the predetermined requirement is the number of the face image data items constituting the face image dataset, a resolution of the face image data items, an attribute to be subjected to labeling, or statistics of attribute values of attributes of the face image data items (Paragraphs [0097]-[0098]: identity obfuscation manager 200 include a similarity score determiner 280 that is generally responsible for determining a similarity score for an updated input image. A similarity score is a metric measuring the degree of similarity or the degree of difference between the updated input image and the original input image. In some embodiments, the similarity score is utilized to ensure that some aesthetics of the input image are retained in the updated user input image. In these cases, the similarity score determined for an updated input image is compared to a predetermined minimum threshold similarity score, and if the similarity score for the updated input image does not meet the minimum threshold, a user is provided a notification indicating that identity obfuscation will change the aesthetics of the user's input image. Additionally or alternatively, the similarity score may be utilized to ensure that the updated input image is not too similar to the input image so that the identity is not properly protected. As such, in these instances, the similarity score determined for an updated input image is compared to a predetermined maximum threshold similarity score, and if the similarity score for the updated input image does not meet the maximum threshold, a user is provided a notification indicating that facial identity could not be protected. Where the updated input image does not meet the maximum threshold score, the updated input image may either be presented with the notification or may not be presented at all. In some embodiment, the determined similarity score is compared to both a maximum score and a minimum score to determine whether the updated input image both sufficiently obfuscates the individual's identity and preserves the aesthetics of the original input image. Further, in some embodiments, the similarity score is computed and provided as feedback to other components of identity obfuscation manager 200 to help improve the functioning of identity obfuscation manager 200…Exemplary embodiments of similarity score determiner 280 determines the similarity score by computing a structural similarity (SSIM) index value. SSIM index is a method in image processing used for predicting the perceived quality of an image. The SSIM index value is calculated on various windows of an image. The measure between two windows x and y of common size (N×N) is determined).
Regarding claim 12, Tagra discloses the data creation method according to claim 1, wherein, by converting the source image to a numerical value, the data creation device creates the creation parameter according to the source image (Fig. 4A; Fig. 5; Paragraph [0055]: Process 400 includes, at preliminary stage 410, resizing an input image 412. Resizing may include rescaling input image 412 into different resolutions to form an image pyramid 414, which is a set of images corresponding to the input image 412 but having different resolutions. Image pyramid 414 is input into MTCNN such that each image of image pyramid 414 is run through the three network stages of MTCNN. A first stage 420 comprises a proposal network (P-Net), which is a shallow convolutional neural network (CNN) that identifies candidates for the most likely facial regions within the input image 412. Candidates may be calibrated using estimated bounding box regression, and certain overlapped candidates may be merged using non-maximum suppression (NMS). Candidates identified in the first stage 420 are input into a second stage 430, which comprises a refine network (R-Net). R-Net is a CNN that filters out the false candidates. The second stage 430 may also include NMS and bounding box regression to calibrate the determined bounding boxes. A third stage 440 comprises an output network (O-Net), which is a CNN that detects particular details within the facial region candidates remaining after the second stage 430. In exemplary aspects, O-Net outputs a final bounding box 442 for a face and select facial landmarks 444. The facial landmarks output by O-Net are landmarks for the eyes, nose and mouth; Paragraph [0066]: a set of similar facial images are selected from the reference facial images, where the selected facial images are the top n reference facial images that have the most similarity with the base facial image. For embodiments utilizing Euclidean distance, the reference facial images with the smallest distance are selected as most similar to the base facial images. In some embodiments, the number of images selected for the set of similar images (n) may be a value within a range of two and ten. For example, the set of similar facial images may consist of six reference facial images that have been determined to be most similar to the base facial image. FIG. 5 depicts an example set of similar facial images 510 that may be identified for the example base facial image 512).
Regarding claim 13, Tagra discloses the data creation method according to claim 1, wherein the data creation device creates the face image data items by using a GAN or a VAE according to the input creation parameters (Paragraph [0039]: Some existing techniques for protecting facial identity involve superimposing a stock image of a person's face over a user's uploaded image to change the face in the user uploaded image. While this technique may protect the user's identity, it does so at the expense of the privacy of the actual individual in the other image used. Some existing methods utilizing Generative Adversarial Networks (GANs) synthesize a new face instead of using a stock image).
Regarding claim 14, the limitations of this claim substantially correspond to the limitations of claim 1; thus they are rejected on similar grounds.
Regarding claim 15, the limitations of this claim substantially correspond to the limitations of claim 1; thus they are rejected on similar grounds.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MATTHEW D SALVUCCI whose telephone number is (571)270-5748. The examiner can normally be reached M-F: 7:30-4:00PT.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, XIAO WU can be reached at (571) 272-7761. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MATTHEW SALVUCCI/Primary Examiner, Art Unit 2613
Prosecution Insights