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 . 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 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.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 03/18/2026 has been entered.
Claims 1, 3-7, 9-11, 14-17, 19-22 are pending.
Claims 2, 8, 12, 13 and 18 are cancelled.
Claims 1, 10, 11, 21 and 22 are amended.
Claim Interpretation
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.
Claim 21 recites “means” plus “function” as below:
“a means for providing a neural transfer image…”, “a means for generating a puzzle…”, “a means for generating a CAPTCH…”, “a means for performing the CAPTCH…”, “a means for…linking the missing block…”, “a means for moving the missing block…”, “ a means for sending a CAPTCHA answer…”, “a means for validating the CAPTCH…”, “a means for selecting…”, “a means for identifying multiple locations…”, “a means for carving a hollow block…”, “a means for selecting a missing block…”. Therefore Claim 21 invokes 112(f).
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 § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 3-4, 7, 9-12, 14, 15, 17 and 19-22 are rejected under 35 U.S.C. 103 as being unpatentable over Aleksandrovich et al (US 2012/0323700 A1) (hereinafter “Aleksandrovich”) and in view of Kubendran (U.S. PGPub. No. 2021/0142454) (hereinafter “Kubendran”) and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”; and in further view of Conti et al (U. S. Pat. No 10,387,645 B2) (Hereinafter “Conti”) and Isaacs (U. S. Pat. No. 8,671,058 B1) (hereinafter “Isaacs”); further in view of Okada (U. S. PGPub. No. 2016/0129339 A1) (hereinafter “Okada”).
Regarding Claim 1, Aleksandrovich teaches:
A system comprising (Aleksandrovich: [0031], a system for providing CAPTCHA security to websites comprising):
one or more processors (Aleksandrovich: [0031], 1) a client device having a processor and a storage medium including machine readable instructions that when executed by a client cause the client device to load a webpage, including a CAPTCHA challenge; (2) a CAPTCHA server having a processor and a storage medium including machine readable instructions that when executed are capable of performing the steps);
and memory storing instructions that, when executed by the one or more processors, cause the system to perform (Aleksandrovich: [0031], (3) a secured website server having a processor and a computer readable storage medium including machine readable instructions that when executed perform the steps of):
a CAPTCHA server coupled to the CAPTCHA puzzle generator, wherein, in operation (Aleksandrovich: [0016] The present invention is generally directed to a method for remote verification of human interaction comprising the steps of receiving a request for a CAPTCHA challenge with a CAPTCHA server; generating the CAPTCHA challenge; generating a unique identifier related to the CAPTCHA challenge; and storing a CAPTCHA challenge solution on a CAPTCHA server):
to the CAPTCHA puzzle generator (Aleksandrovich: [0006], provides for One common way to differentiate a human from a computer is by a test known as a "Turing test." When a computer program is able to generate the Turing test (=CAPTCHA generator) and evaluate the results, it is typically known as a CAPTCHA (completely automated public test to tell computer and humans apart) program. In addition to the general desire not have certain portions, functions, areas, content or privileges of a website freely available to automated systems, many websites use CAPTCHA programs to prevent attacks by malicious programs, including those that are designed to disrupt service on a large scale. [0020] The CAPTCHA challenge may include an image or graphical representation, which may instruct the client on how to manipulate the graphical elements and wherein the image is capable of being manipulated to match the graphical representation of the CAPTCHA challenge solution. [0021] The CAPTCHA challenge generally includes a graphical representation and graphical elements and at least one of the graphical representation and graphical elements may be distorted such that the graphical elements created from the graphical representation are no longer identical, and when a client solution is assembled, it includes differences between the assembled graphical elements and the graphical representation. The challenge solution stored on the CAPTCHA server includes the graphical coordinates of the graphical elements, such as the graphical coordinates of the assembled graphical elements when the match the graphical representation or desired solution).
the CAPTCHA puzzle generator generates a puzzle by associating multiple hollow blocks (Aleksandrovich: [0020]: provides for the edges of the graphical elements may intentionally overlap, include spaces or other misalignments. [0031], provides for generating a CAPTCHA challenge having a graphical representation and at least one graphical element that is capable of being rearranged. [0063] The CAPTCHA challenge 24 is typically presented to the user of the website within a specified area on the website page, such as in the exemplary box 10. Although the CAPTCHA challenge 24 is illustrated as being presented in a box 10, it may be easily displayed on the webpage without the box 10 or in a variety of other settings. As used herein the terms box, area and space occupied by the moveable pieces of the challenge may be used interchangeably. The box 10 generally contains a challenge 24, such as a puzzle, having a graphical representation 22 of the desired solution, and at least one graphical element 20 requiring manipulation or assembly, such as the illustrated puzzle pieces in in FIGS. 1-3).
the CAPTCHA server, which includes a CAPTCHA generator that generates a CAPTCHA comprised of (Aleksandrovich: [0016], generating the CAPTCHA challenge. [0019], provides for the CAPTCHA challenge is configured to include a graphical representation and graphical elements which are capable of being rearranged to match the graphical representation. The graphical representation is used to generate graphical elements and wherein at least one of the graphical representations and the graphical elements (=Correct hollow block) are manipulated by at least one process of enlargement, rotation, shifting, or overlaying on different backgrounds. The graphical elements include edges which when arranged to match the graphical representation, may not be aligned. For example, gaps, overlays and other variances may be intentionally added.),
a missing block (Aleksandrovich: [0071] FIGS. 4-5 and 9-12 provide other types of assembly puzzles and are provided as only exemplary style puzzles. In FIGS. 4 and 5, the client is presented with partial images (=partial image contains missing hollow block (=blank space) shown in FIG. 7) on the background and then assembles or manipulates the graphical elements 20, such as the various butterflies, depending upon shape and color).
and an incorrect hollow block (Aleksandrovich: [0064]: To increase the difficulty of the challenge 24 for automated systems, a background 12, such as additional completed butterflies or portions of butterflies occurring in the background 12 but not part of the graphical elements 20, may be included (=incorrect hollow block)
performs the CAPTCHA in association with a CAPTCHA client device (Aleksandrovich: [0062], provides for to obtain access to the desired website area, functionality or content, the client or user must solve a CAPTCHA challenge. Exemplary CAPTCHA challenges of the present invention are illustrated in FIGS. 1-21).
Alesandrovich does not explicitly disclose:
neural style transfer engine for combining the content image and the style image into the neural style transferred image.
However, in an analogous art, Kubendran teaches:
(Kubendran: [0018] The system includes several modules that facilitate the generation of graphic design patterns. In FIG. 2, a block diagram shows a system 200 according to an example embodiment. The system 200 takes two input images, a content image 202 and a style image 204. The content image 202 could be a silhouette and/or a picture from which a silhouette can be extracted, e.g., via image to silhouette convertor 210. The style image 204 could be a pattern, picture, or any other image that has some desirable stylistic properties. Optional image-to-silhouette 210 module and procedural generation 212 module can be selected if the input images are pictures (e.g., photographic images) instead of a silhouette and a pattern. These are input to a neural style transfer processor 2056 that processes the input images 202, 204 to produce output images 206, 208 (=the neural style transferred images)).
(Kubendran: [0018] The system includes several modules that facilitate the generation of graphic design patterns. In FIG. 2, a block diagram shows a system 200 according to an example embodiment. The system 200 takes two input images, a content image 202 and a style image 204. The content image 202 could be a silhouette and/or a picture from which a silhouette can be extracted, e.g., via image to silhouette convertor 210. The style image 204 could be a pattern, picture, or any other image that has some desirable stylistic properties. Optional image-to-silhouette 210 module and procedural generation 212 module can be selected if the input images are pictures (e.g., photographic images) instead of a silhouette and a pattern. These are input to a neural style transfer processor 2056 that processes the input images 202, 204 to produce output images 206, 208 (=the neural style transferred images))
(Kubendran: [0018] The system includes several modules that facilitate the generation of graphic design patterns. In FIG. 2, a block diagram shows a system 200 according to an example embodiment. The system 200 takes two input images, a content image 202 and a style image 204. The content image 202 could be a silhouette and/or a picture from which a silhouette can be extracted, e.g., via image to silhouette convertor 210. The style image 204 could be a pattern, picture, or any other image that has some desirable stylistic properties. Optional image-to-silhouette 210 module and procedural generation 212 module can be selected if the input images are pictures (e.g., photographic images) instead of a silhouette and a pattern. These are input to a neural style transfer processor 2056 that processes the input images 202, 204 to produce output images 206, 208 (=the neural style transferred images))
(Kubendran: [0037], A neural network 930 (e.g., a deep, convolutional neural network) uses the silhouette image to produce content feature layers and uses the style image to produce pattern feature layers. The content feature layers and pattern feature layers can be stored in the memory 904 and or persistent storage 906 (=content images and style image datastore)).
an image selection engine for selecting a content image from the content image datastore and a style image from the style image datastore (Kubendran: [0034] & [0037]In FIG. 8, a flowchart shows a method according to an example embodiment. The method involves inputting 800 a silhouette image to a deep convolutional neural network to produce content feature layers. A style image is input 801 to the deep convolutional neural network to produce pattern feature layers. The content feature layers and the pattern feature layers from the deep convolutional neural network are combined 802 to obtain an output image. The output image includes an abstraction of the style image within confines of the silhouette image. The output image is utilized 803 in a graphic design product),
and a neural style transfer engine for combining the content image and the style image into the neural style transferred image (Kubendran: [0018] The system includes several modules that facilitate the generation of graphic design patterns. In FIG. 2, a block diagram shows a system 200 according to an example embodiment. The system 200 takes two input images, a content image 202 and a style image 204. The content image 202 could be a silhouette and/or a picture from which a silhouette can be extracted, e.g., via image to silhouette convertor 210. The style image 204 could be a pattern, picture, or any other image that has some desirable stylistic properties. Optional image-to-silhouette 210 module and procedural generation 212 module can be selected if the input images are pictures (e.g., photographic images) instead of a silhouette and a pattern. These are input to a neural style transfer processor 2056 that processes the input images 202, 204 to produce output images 206, 208)
the neural style transferred image (Kubendran: [0018] The system includes several modules that facilitate the generation of graphic design patterns. In FIG. 2, a block diagram shows a system 200 according to an example embodiment. The system 200 takes two input images, a content image 202 and a style image 204. The content image 202 could be a silhouette and/or a picture from which a silhouette can be extracted, e.g., via image to silhouette convertor 210. The style image 204 could be a pattern, picture, or any other image that has some desirable stylistic properties. Optional image-to-silhouette 210 module and procedural generation 212 module can be selected if the input images are pictures (e.g., photographic images) instead of a silhouette and a pattern. These are input to a neural style transfer processor 2056 that processes the input images 202, 204 to produce output images 206, 208 (=the neural style transferred images)).
It would be obvious to a person having ordinary skill in the art, before the effective filing date of the invention, to modify Aleksandrovich’s method of generating, performing and solving CAPTCHA puzzle by applying Kubendran’s method of combining content image and style image in order to produce a neural style transfer image. The motivation is to create an automated graphic design (Kubendran: [0015]).
The combination of Aleksandrovich in view of in view of Kubendran does not explicitly teaches:
and applies the puzzle
However, in an analogous art, “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” teaches:
and applies the puzzle (Image-Based CAPTCHA’s: [abstract], In this study, the authors apply neural style transfer to enhance the security for CAPTCHA design. [Page no. 521, Col 2, section 4.1, para 3], The final CAPTCHA image is generated (=creating CAPTCHA) via a combination of background image and icons with random locations. Both the background image and the icons are generated using the same style. In this case, the texture of the foreground is very similar to that of background, hence, it is more difficult for attackers to segment the target icons);
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of in view of Kubendran by applying the well-known technique as disclosed by “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” of creating CAPTCHA image in order to make difficult for attacker to segment the target image. The motivation is to make the challenge more difficult to prevent websites from malicious attack (“Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” : [Abstract]).
Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” does not explicitly teach:
when the CAPTCHA is displayed to a user of the CAPTCHA client device, the missing block is linked to a header of a range slider;
when the header is dragged to a correct location, the missing block is moved to an inline location of the correct hollow block on the image, wherein movement of the missing block is coordinated with movement of the header;
wherein the CAPTCHA client device sends a CAPTCHA answer that includes a slider distance position representing distance the header is moved;
and validating the CAPTCHA based on the slider distance position representing the distance the header is moved.
However, in analogous art, Conti teaches:
when the CAPTCHA is displayed to a user of the CAPTCHA client device, the missing block is linked to a header of a range slider (Conti: [Col 5, lines 40-44], (15) With the term “Sensibility” it is meant to refer to a parameter which indicates the intensity of dislocation of each geometric shape with respect to the cursor movement FIG. 1 shows a flow diagram of the method according to a suggested embodiment of the invention);
when the header is dragged to a correct location, the missing block is moved to an inline location of the correct hollow block on the image, wherein movement of the missing block is coordinated with movement of the header (Conti: [Col 6, lines 56-62], (27) Once the method visualized the distorted image on the display of an electronic terminal, and generated the coordinates of the solution position, the method provides to detect (step 105) the movement of a cursor within the test area. The cursor can be of any type, for example a pointer of a mouse, a pointer of an optic pen, the result of the pressure of a finger on a touch screen);
wherein the CAPTCHA client device sends a CAPTCHA answer that includes a slider distance position representing distance the header is moved (Conti: [Col 7, lines 54-60], (34) So the method detects the final position of the cursor (step 108), where the final position is the position of the cursor when the user input the control signal. In particular, every time the user moves the cursor during the test, the method provides to use the coordinates of the cursor (cur.sub.x, cur.sub.y) in the test area and uses them to compute, moment by moment, the position of each image portion using the following formulas: x.sup.i=m.sub.xx.sup.i.Math.cur.sub.x+m.sub.xy.sup.i.Math.cur.sub.y+C.sub.x.sup.i
y.sup.i=m.sub.yy.sup.i.Math.cur.sub.y+m.sub.yx.sup.i.Math.cur.sub.x+C.sub.y.sup.i);
and validating the CAPTCHA based on the slider distance position representing the distance the header is moved (Conti: [Col 7. Lines 66-76 to col 8, lines 1-25], (35) The user stops the movement of the cursor when the user believes that the cursor is in the final position (cur.sub.x.sup.f, cur.sub.y.sup.f) where the user recognizes the distribution of image portions inside the test area to be the original image. So such method provides that when the client terminal detects the final position of the cursor, it transmits the coordinates of the final position of the cursor to the server terminal, which accepts the coordinates of the final position (cur.sub.x.sup.f, cur.sub.y.sup.f). Subsequently, the server terminal compares such coordinates with the coordinates of the solution position (sol.sub.x, sol.sub.y) through a script therein implemented. This comparison occurs by comparing the euclidean distance between the final position and the solution position, and a predetermined threshold of tolerance. If such difference is less than the tolerance threshold, the method considers that the interaction with the electronic terminal is accomplished by a human, and therefore the user has passed the test).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed by Conti’s method of comparing and calculating the euclidean distance between the final position and the solution position, and therefore the user has passed the test. The motivation is to improve the ease of usability, level of security, the efficiency of a test for recognizing if the user of an electronic terminal is a human or a robot (Conti: [Col 2, lines 54-64]).
Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and does not explicitly teach:
selecting a hollow shape, wherein the hollow shape is procedurally generated;
identifying multiple locations in the neural style transferred image for hollow block placement, wherein at least two of the multiple locations are fixed locations apart from each other;
carving a hollow block at each of the multiple locations in the neural style transferred image;
and selecting a missing block that matches a correct hollow block at one of the fixed locations in the neural style transferred image.
However, in a an analogous art, Isaacs teaches:
selecting a hollow shape, wherein the hollow shape is procedurally generated (Isaacs: [Col 13, Col 25-32], (80) The selection process for choosing the correct puzzle piece is based on puzzle shape and the image contained on the piece (S667). The puzzle piece options provided for the user to select from could include identical shapes but only one would have the correct image to complete the puzzle. Similarly the puzzle pieces provided for selection could all have the correct image but be of different shapes, with only one piece have the correct shape to complete the puzzle).
identifying multiple locations in the neural style transferred image for hollow block placement, wherein at least two of the multiple locations are fixed locations apart from each other (Isaacs : [FIG. 24], 816, 817 are the two blocks (only one is correct block) and the white space in 819 is the missing piece of the image. [Col 13, lines 33-45], (81) The user chooses a puzzle piece believed to be the correct missing piece (S669). The choosing action can be for example performed by clicking on the piece with the cursor controlled by the mouse or other pointing device of the user interface. If the user is viewing the puzzle captcha puzzle on a user interface with a touch sensitive screen, the choosing action can be performed by user tapping the correct puzzle piece with their finger or other device such as a stylus. If the user is using a mobile phone or other viewing device as a user interface where a pointing device is not present but a joystick or toggle control is supplied, the choosing action can be performed by operator navigates to the correct puzzle piece and selects the piece);
and selecting a missing block that matches a correct hollow block of the procedurally generated hollow blocks at one of the fixed locations in the (Isaacs: [Col 11, lines 66-67-Col 12, lines 1-7], (73) As already mentioned above, in one implementation, the puzzle is a jigsaw puzzle (see for example FIG. 20). The Puzzle Captcha system places substantially equal weight on color, shape and pattern to create a puzzle that is simple to solve for a human but extremely difficult to solve for a computer. As a general overview, an incomplete jigsaw puzzle is randomly generated (see for example FIG. 22). The incomplete jigsaw puzzle has one or more pieces missing. Solution pieces are also provided (see for example FIGS. 23 & 24). These provided solution pieces include the piece missing from the provided incomplete jigsaw puzzle. Solving skills required to solve the puzzle include (i) visually matching provided puzzle pieces (solution pieces) with the missing piece that forms part of those pieces that have been assembled; (ii) choosing the correct piece requires the user to interpret shape and surface patterns. The jigsaw puzzle solution pieces may include identical pieces with the correct solution shape but only one piece will have the correct image section).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed by Isaacs’s method of completing the CAPTCHA puzzle by choosing the correct piece from the multiple pieces in order to complete the CAPTCHA puzzle. The motivation is to provide an improved method of generating a Completely Automated Public Test to tell Computers and Humans Apart (CAPTCHA), and more particularly to a method that can be used as an entry point to online websites or protected sections, pages or links of websites (Isaacs: [Col 2, lines20-25]).
The Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Isaac does not explicitly disclose:
carving a hollow block at each of the multiple locations in the neural style transferred image, wherein a shape of the hollow blocks corresponds to the procedurally generated hollow shape, and wherein each of the carved hollow blocks is translucent;
However, in an analogous art, Okada disclose:
carving a hollow block at each of the multiple locations in the neural style transferred image (Okada: [0078], the puzzle generation unit 21 may cut out graphics of predetermined shape, cut out (=carving) graphics of shapes selected at random from among a plurality of shapes, or the like. [0148], For example, in FIG. 9, an advertisement for a travel company has been cut out (=carving) at three locations. The information processing device 1 displays a partial image 91, objects 92-94, and blank areas 95-97 (=hollow block). Preferably, when in this state, the decision unit 19 has detected that a user has dragged the object 92 with a finger and arranged it in the blank area 95, dragged the object 93 with a finger and arranged it in the blank area 96, and dragged the object 94 with a finger and arranged it in the blank area 97, the execution unit 20, using predetermined executable information, will execute a predetermined action), wherein a shape of the hollow blocks corresponds to the procedurally generated hollow shape (Okad: [0078], , the puzzle generation unit 21 may cut out graphics of predetermined shape, cut out graphics of shapes selected at random from among a plurality of shapes, or the like. That is, the specific method of acquiring graphics is of no particular concern. Herein, a graphic refers to an object), and wherein each of the carved hollow blocks is translucent (Okada: [0059], A blank area (=hollow block) may be an area bounded by broken lines, solid lines, or the like, or a transparent area. In this case, the blank area will be an area that makes up part of a background image, and as a design may be considered as being continuous with the background image.)
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed by Okada’s method of generating cut-out graphics of predetermined shape, cut out graphics of shapes selected at random from among a plurality of shapes and making blank area as transparent. The motivation is generating puzzles by cut out shapes of an image, blank area, wherein blank area is transparent or it will be part of background image in order to make it difficult for attacker to identify puzzle pieces.
Regarding Claim 3, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”; and in further view of Conti, Isaacs and Okada teaches:
The system of claim 1 (see rejection of claim 1 above),
A slider block selection engine (Conti: [Col 6, lines 63-67 to Col 7, lines 1-3], (28) According to the movement of the cursor (in FIG. 5, the cursor is represented by a white arrow and indicated with the references 505 and 506) in the test area, the method provides to move (step 106) each image portion inside the test area, which generates the different images 503 and 504, as reported by FIGS. 5d and 5e that show the evolution of the distorted image 502 when the cursor gets more and more close to the solution position)
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed Conti’s method of movement of the cursor in the test area to move each image portion inside the test area to get the solution position. The motivation is to improve the ease of usability, level of security, the efficiency of a test for recognizing if the user of an electronic terminal is a human or a robot (Conti: [Col 2, lines 54-64]).
The combination of Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and in further view of Conti, Isaacs and Okada does not explicitly disclose:
However, in an analogous art, Okada teaches:
(Okada: [0060] The blank area storage (=hollow shape datastore) means 12 is able to store one or more items of blank area information. Blank area information contains blank area position information, which is information that relates to the position of a blank area within a partial image),
a hollow shape selection engine, a hollow block placement engine, a hollow block carving engine, (Okada: [0078] The puzzle generation unit (=hollow block carving engine) 21 acquires puzzle information. Puzzle information contains one or more objects, a partial image, and blank area information. The puzzle generation unit (=acting as a hollow block carving engine) 21 reads out a full image from the full image storage unit 10, cuts out a portion of the full image in question, and acquires puzzle information that contains one or more objects which are cut-out sections (=hollow block), a partial image from which one or more objects has been cut out, and blank area information composed of information about a blank area which is an area where an object has been cut out. Here, the puzzle generation unit 21 may cut out graphics of predetermined shape, cut out graphics of shapes selected at random from among a plurality of shapes, or the like. [0148], provides for example, in FIG. 9, an advertisement for a travel company has been cut out at three locations).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”, Conti and Isaacs by applying the well-known technique as disclosed by Okada’s method of the generating CAPTCHA puzzle pieces by carving of a full or partial image, in order to generate one or more pieces of an image. The motivation is to performs a predetermined action in cases where one or more objects have been moved to a prescribed blank area on a screen [Okada: [0001]).
Regarding Claim 4, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs and Okada teaches:
The system of claim 1 (see rejection of claim 1 above),
The above cited combination of Aleksandrovich in view of Kubendran teaches:
(Kubendran: [0018] The system includes several modules that facilitate the generation of graphic design patterns. In FIG. 2, a block diagram shows a system 200 according to an example embodiment. The system 200 takes two input images, a content image 202 and a style image 204. The content image 202 could be a silhouette and/or a picture from which a silhouette can be extracted, e.g., via image to silhouette convertor 210. The style image 204 could be a pattern, picture, or any other image that has some desirable stylistic properties. Optional image-to-silhouette 210 module and procedural generation 212 module can be selected if the input images are pictures (e.g., photographic images) instead of a silhouette and a pattern. These are input to a neural style transfer processor 2056 that processes the input images 202, 204 to produce output images 206, 208 (=the neural style transferred images)).
The above cited combination of Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and in further view of Conti and Isaacs, does not explicitly disclose:
However, Okada teaches:
(Okada: [0148], provides for example, in FIG. 9, an advertisement for a travel company has been cut out at three locations (=multiple options). The information processing device 1 displays a partial image 91, objects 92-94, and blank areas 95-97 [0155] Specific Example 4 describes a case in which a single object is able to be arranged in multiple blank areas, and different actions are executed depending on the blank area in which [the object] is arranged)).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs by applying the well-known technique as disclosed by Okada’s method of generating multiple pieces from a full image in order to make difficult to solve the puzzle CAPTCHA by automated system. The motivation is to performs a predetermined action in cases where one or more objects have been moved to a prescribed blank area on a screen [Okada: [0001]).
Regarding Claim 7, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”; and in further view of Conti, Isaacs and Okada teaches:
The system of claim 1 (see rejection of claim 1 above),
(“Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”: [Page 520, Col 2, section 3.1, line 3-6], provides for the reconstruction process includes two steps: style reconstruction and content reconstruction. For humans, the content of an image is the global structure, whereas the style of an image involves the color and local structures) with feed-forward passes (“Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”: [page 3, Col 1, section 3.2, 5-9], To accelerate the iterative optimisation process, Ulyanov et al. [34] and Johnson et al. [18] combined the benefits of feed-forward image transformation tasks and optimisation-based methods for fast neural style transfer tasks and achieved thousands of times).
It would be obvious to a person having ordinary skill in the art, before the effective filing date of the invention, to modify Aleksandrovich’s in view of Kubendran by applying “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” ’s method of reconstruction process of an image in order to transform an image. The motivation is to generate new and unique image.
Regarding Claim 9, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs, Okada teaches:
The system of claim 1, (see rejection of claim 1 above),
The above cited combination of Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs does not explicitly disclose:
wherein the hollow block is debossed in nature.
However, Okada teaches:
wherein the hollow block is debossed in nature (Okada: [0018]: provides a puzzle generation unit that cuts out a part of a full image, and acquires puzzle information that includes one or more objects that are cut-out sections, a partial image from which one or more objects have been cut out, and blank area information containing information about a blank area (=debossed or empty area) that is a cut-out area).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs by applying the well-known technique as disclosed by Okada’s method of creating cutouts of an image to create a blank area (=empty space or debossed portion) of an image. The motivation is to performs a predetermined action in cases where one or more objects have been moved to a prescribed blank area on a screen [Okada: [0001]).
Regarding Claim 10, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and in further view of Conti and Isaacs, Okada teaches:
The system of claim 1 (see rejection of claim 1 above),
wherein the CAPTCHA server includes a token generator, a CAPTCHA queue, an object datastore, a token validator, and a CAPTCHA validator (Aleksandrovich: [0076]: provide for as discussed below, the CAPTCHA server 34 then verifies, authenticates or matches the client solution 28 to a stored CAPTCHA challenge solution 26 before access is granted to the client),
(Conti: [Col 7. Lines 66-76 to col 8, lines 1-25], (35) The user stops the movement of the cursor when the user believes that the cursor is in the final position (cur.sub.x.sup.f, cur.sub.y.sup.f) where the user recognizes the distribution of image portions inside the test area to be the original image. So such method provides that when the client terminal detects the final position of the cursor, it transmits the coordinates of the final position of the cursor to the server terminal, which accepts the coordinates of the final position (cur.sub.x.sup.f, cur.sub.y.sup.f). Subsequently, the server terminal compares such coordinates with the coordinates of the solution position (sol.sub.x, sol.sub.y) through a script therein implemented. This comparison occurs by comparing the euclidean distance between the final position and the solution position, and a predetermined threshold of tolerance. If such difference is less than the tolerance threshold, the method considers that the interaction with the electronic terminal is accomplished by a human, and therefore the user has passed the test)).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed by Conti’s method of comparing and calculating the euclidean distance between the final position and the solution position, and therefore the user has passed the test. The motivation is to improve the ease of usability, level of security, the efficiency of a test for recognizing if the user of an electronic terminal is a human or a robot (Conti: [Col 2, lines 54-64]).
Regarding claim 11, this claim contains identical limitations found within that of claim 1
above albeit directed to a different statutory category (method medium). For this reason, the same grounds of rejection are applied to claim 11.
Regarding Claim 14, this claim contains identical limitations found within that of claim 4 above albeit directed to a different statutory category (method medium). For this reason, the same grounds of rejection are applied to claim 14.
Regarding Claim 15, this claim contains identical limitations found within that of claim 5 below albeit directed to a different statutory category (method medium). For this reason, the same grounds of rejection are applied to claim 15.
Regarding Claim 16, this claim contains identical limitations found within that of claim 6 below albeit directed to a different statutory category (method medium). For this reason, the same grounds of rejection are applied to claim 16.
Regarding Claim 17, this claim contains identical limitations found within that of claim 7 above albeit directed to a different statutory category (method medium). For this reason, the same grounds of rejection are applied to claim 17.
Regarding Claim 19, this claim contains identical limitations found within that of claim 9 above albeit directed to a different statutory category (method medium). For this reason, the same grounds of rejection are applied to claim 19.
Regarding Claim 20, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and in further view of Conti and Isaacs, Okada teaches:
The method of claim 11 (see rejection of claim 11 above),
generating a token (Aleksandrovich: [0078]: generates an MD5 signature (=a token) to protect all transferred data between the secured website, the client's web browser, and the CAPTCHA server 34);
queuing a plurality of CAPTCHAs, including the CAPTCHA (Aleksandrovich: [0074], the ability of automated systems to keep up with the ever-increasing number of CAPTCHAs (=queuing CAPTCHAS), given the wide variety of types of images that would be used, is limited. In creating the CAPTCHA type puzzle, a marketer or website owner would submit a copy of an image to the CAPTCHA system wherein the CAPTCHA system would automatically enter and upload the image into the database and then create the desired puzzle);
maintaining an object datastore that includes the CAPTCHA, a CAPTCHA answer, the token, and a CAPTCHA client UID stored in association with one another (Aleksandrovich: [0091] The CAPTCHA server 34 accepts the MD5 signature (=the token) and verifies that such signature was generated by the secured website server 32. [0096] The secured website server 32 then sends a request to the CAPTCHA server 34 and such request includes the unique identifier (=CAPTCHA client UID) 40 generated in step 135 of FIG. 23. In some instances, it may also include the client solution (=a CAPTCHA answer). In step 200 of FIG. 25 the CAPTCHA server 34 accepts the unique identifier 40 generated in step 135 of FIG. 23, and using this unique identifier 40, the CAPTCHA server 34 then locates the CAPTCHA verification result according to its unique identifier 40 in its internal database (the verification result is stored in the database);
validating the CAPTCHA (Aleksandrovich: [0091], provides for the CAPTCHA server 34 accepts the MD5 signature and verifies that such signature was generated by the secured website server 32. FIG.24 illustrates in steps 165 and 170 that the CAPTCHA server 34 may confirm that the client solution constituting a proposed solution to a given CAPTCHA challenge 24 submitted by the client is correct by comparing, verifying, matching or authenticating the client solution 28 submitted by the client against the stored CAPTCHA challenge solution 26. More specifically, the CAPTCHA solution stored on the CAPTCHA server 34 and associated with the unique identifier 40, as described above, are both compared, matched, verified, or authenticated against the client solution (=a CAPTCHA answer) and associated unique identifier 40 sent by the client device);
validating the CAPTCHA client UID by matching it with the CAPTCHA client UID stored in the object datastore (Aleksandrovich: [0091], provides for the CAPTCHA solution stored on the CAPTCHA server 34 and associated with the unique identifier 40, as described above, are both compared, matched, verified, or authenticated against the client solution and associated unique identifier (=UID) sent by the client device. [0096], provides for the secured website server 32 then sends a request to the CAPTCHA server 34 and such request includes the unique identifier 40 generated in step 135 of FIG. 23. In some instances, it may also include the client solution. In step 200 of FIG. 25 the CAPTCHA server 34 accepts the unique identifier (=UID) generated in step 135 of FIG. 23, and using this unique identifier 40, the CAPTCHA server 34 then locates the CAPTCHA verification result according to its unique identifier 40 in its internal database (the verification result is stored in the database)).
validating the token validating a received CAPTCHA answer by checking if(Aleksandrovich: [0091], provides for the CAPTCHA server 34 accepts the MD5 signature and verifies that such signature was generated by the secured website server 32. FIG.24 illustrates in steps 165 and 170 that the CAPTCHA server 34 may confirm that the client solution constituting a proposed solution to a given CAPTCHA challenge 24 submitted by the client is correct by comparing, verifying, matching or authenticating the client solution 28 submitted by the client against the stored CAPTCHA challenge solution 26) in the received CAPTCHA answer is equal to the CAPTCHA answer in the object datastore (Aleksandrovich: [0027], provides for the step of initiating a verification process includes the step of verifying movement of each graphical element of the CAPTCHA challenge from an initial position. [0029], provide for verifying with the client device movement of each graphical element of the CAPTCHA challenge from an initial position (position of graphical element from initial position to the expected position = distance traveled/covered by slider). [0065], provides for FIGS. 3, 10 and 12, the client has moved or manipulated all of the graphical elements 20 or puzzle pieces to the proper position (=equal to the CAPTCHA answer in the datastore/expected position) and perfectly assembled the puzzle such that a verification or submit control 38 may be pressed to check the correctness of the CAPTCHA assembly. [0091], provides for FIG.24 illustrates in steps 165 and 170 that the CAPTCHA server 34 may confirm that the client solution constituting a proposed solution to a given CAPTCHA challenge 24 submitted by the client is correct by comparing, verifying, matching or authenticating the client solution 28 submitted by the client against the stored CAPTCHA challenge solution 26). [0092] While the CAPTCHA server 34 could store the actual graphical solution, such as an image, on the CAPTCHA server 34, it typically saves the coordinates (of the moveable objects) when the CAPTCHA challenge 24 is being formed in the step 130 of FIG. 23)
(Conti: [Col 7. Lines 66-76 to col 8, lines 1-25], (35) The user stops the movement of the cursor when the user believes that the cursor is in the final position (cur.sub.x.sup.f, cur.sub.y.sup.f) where the user recognizes the distribution of image portions inside the test area to be the original image. So such method provides that when the client terminal detects the final position of the cursor, it transmits the coordinates of the final position of the cursor to the server terminal, which accepts the coordinates of the final position (cur.sub.x.sup.f, cur.sub.y.sup.f). Subsequently, the server terminal compares such coordinates with the coordinates of the solution position (sol.sub.x, sol.sub.y) through a script therein implemented. This comparison occurs by comparing the euclidean distance between the final position and the solution position, and a predetermined threshold of tolerance. If such difference is less than the tolerance threshold, the method considers that the interaction with the electronic terminal is accomplished by a human, and therefore the user has passed the test) ;
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed by Conti’s method of comparing and calculating the euclidean distance between the final position and the solution position. The motivation is to improve the ease of usability, level of security, the efficiency of a test for recognizing if the user of an electronic terminal is a human or a robot (Conti: [Col 2, lines 54-64]).
Regarding claim 21, this claim contains identical limitations found within that of claim 1
above albeit directed to a different statutory category (system medium). For this reason, the same grounds of rejection are applied to claim 21.
Regarding Claim 22, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and in further view of Conti and Isaacs teaches:
solving a CAPTCHA on the CAPTCHA client device with a client UID, wherein the missing block is moved by the user by moving the header of the range slider to an inline location of the at least one hollow block on the neural style transferred image (Aleksandrovich: [0068]: the CAPTCHA server 34 operator to vary the amount of variance the graphical elements 20 have in placement, such that the task provided as a challenge 24 is considered to be solved or match a given solution, even if the client has not assembled the entire image or graphical elements 20 precisely. Another benefit to using the illustrated puzzle-based verification system is that many touch screen devices such as smart phones, music players, and tablets can be cumbersome in entering text-based CAPTCHA challenge solutions),
and (Aleksandrovich: [0027], provides for the step of initiating a verification process includes the step of verifying movement of each graphical element of the CAPTCHA challenge from an initial position (=position distance). [0029], provide for verifying with the client device movement of each graphical element of the CAPTCHA challenge from an initial position. [0065], provides for FIGS. 3, 10 and 12, the client has moved or manipulated all of the graphical elements 20 or puzzle pieces to the proper position and perfectly assembled the puzzle such that a verification or submit control 38 may be pressed to check the correctness of the CAPTCHA assembly. [0091], provides for FIG.24 illustrates in steps 165 and 170 that the CAPTCHA server 34 may confirm that the client solution constituting a proposed solution to a given CAPTCHA challenge 24 submitted by the client is correct by comparing, verifying, matching or authenticating the client solution 28 submitted by the client against the stored CAPTCHA challenge solution 26). [0092] While the CAPTCHA server 34 could store the actual graphical solution, such as an image, on the CAPTCHA server 34, it typically saves the coordinates (of the moveable objects) when the CAPTCHA challenge 24 is being formed in the step 130 of FIG. 23));
sending the CAPTCHA answer along with the client UID to a server for validation (Aleksandrovich: [0017] The method may also associate the unique identifier (=Client UID) with the CAPTCHA challenge solution (=CAPTCHA answer), as well as store the unique identifier related to the CAPTCHA challenge and the CAPTCHA challenge solution on the CAPTCHA server. Of course, it is expected that any replies from the client device, including a client solution, will also include the unique identifier).
and wherein the distance to which the header is moved is calculated as slider distance position (Conti: [Col 7. Lines 66-76 to col 8, lines 1-25], (35) The user stops the movement of the cursor when the user believes that the cursor is in the final position (cur.sub.x.sup.f, cur.sub.y.sup.f) where the user recognizes the distribution of image portions inside the test area to be the original image. So such method provides that when the client terminal detects the final position of the cursor, it transmits the coordinates of the final position of the cursor to the server terminal, which accepts the coordinates of the final position (cur.sub.x.sup.f, cur.sub.y.sup.f). Subsequently, the server terminal compares such coordinates with the coordinates of the solution position (sol.sub.x, sol.sub.y) through a script therein implemented. This comparison occurs by comparing the euclidean distance between the final position and the solution position, and a predetermined threshold of tolerance. If such difference is less than the tolerance threshold, the method considers that the interaction with the electronic terminal is accomplished by a human, and therefore the user has passed the test) .
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” by applying the well-known technique as disclosed by Conti’s method of comparing and calculating the euclidean distance between the final position and the solution position. The motivation is to improve the ease of usability, level of security, the efficiency of a test for recognizing if the user of an electronic terminal is a human or a robot (Conti: [Col 2, lines 54-64])
Claim(s) 5-7 are rejected under 35 U.S.C. 103 as being unpatentable over Aleksandrovich et al (US 2012/0323700 A1) in view of Kubendran (U.S. PGPub. No. 2021/0142454) (hereinafter “Kubendran”), and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”, Conti et al (U. S. Pat. No 10,387,645 B2) (Hereinafter “Conti”) and Isaacs (U. S. Pat. No. 8,671,058 B1) (hereinafter “Isaacs”); and further in view of “Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review".
Regarding Claim 5, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs teaches:
The system of claim 1 (see rejection of claim 1 above),
The above cited combination of Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and Conti, Isaacs does not explicitly disclose:
wherein the neural style transfer engine includes a 5- layer of Visual Geometry Group (VGG)-19 encoder and decoder with Rectified Linear Unit (ReLU) as an activation function.
However, in an analogous art, “Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review” teaches:
wherein the neural style transfer engine includes a 5- layer of Visual Geometry Group (VGG)-19 encoder and decoder with Rectified Linear Unit (ReLU) as an activation function (“Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review”: [Page 3373, Col 2, para 5, line 4-9]: To speed up the semantic stylisation process, Lu et al. [68] propose to optimize the objective in the feature space, instead of in the pixel space. They first forward the content and style images through a pre-trained VGG encoder to get the corresponding content and style features. [Page no. 3372, Col 2, para2, lines 3-13] To address this limitation, Li et al. [57] propose to model the problem of neural style transfer as a linear transformation between the encoded content features and the learned transformation matrix. They replace the transformation matrix computations between the encoder and decoder in [56] with feed-forward networks, which are trained to directly output the desired transformation matrix. The result of linear multiplication between this learned transformation matrix and the encoded content features is forwarded into the decoder to get the corresponding stylised result).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”, and Conti, Isaacs by applying the well-known technique as disclosed by “Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review” ’s method of using ReLU as a decoder in order to increase the width, height of the input layer and VGG-19 as a encoder in order to increase the neural network’s performance .
Regarding Claim 6, Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”, and Conti, Isaacs teaches:
The system of claim 1 (see rejection of claim 1 above),
The above cited combination of Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer”, and in further view of Conti, Isaacs does not explicitly disclose:
wherein the neural style transfer engine has style- agnostic generation ability with marginally compromised visual quality and execution efficiency.
However, “Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review” teaches:
wherein the neural style transfer engine has style- agnostic generation ability with marginally compromised visual quality and execution efficiency (“Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review”: [Page 3366, Fig 1, Page 3375, Fig 3 and page 3376, Fig 4], provides for the qualitative evaluation using NST).
A person having ordinary skill in the art, before the effective filing date of the invention, would have found it obvious to modify Aleksandrovich in view of Kubendran and “Cheng, Z., Gao, H., Liu, Z., Wu, H., Zi, Y. and Pei, G. (2019), Image-Based CAPTCHAs based on neural style transfer” and in further view of Conti and Isaacs by applying the well-known technique as disclosed by “Y. Jing, Y. Yang, Z. Feng, J. Ye, Y. Yu and M. Song, "Neural Style Transfer: A Review” ’s method of generating a new image in order to generate optimize, transfer an image into a new image in agnostic manner.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Refer to PTO-892, Notice of References Cited for a listing of analogous art.
Gilbert et al. (U. S. Pat. No. 5,769,418): A puzzle has a transparent top layer and a bottom layer which can be seen through the top layer. The top layer of the puzzle is provided with a top image and the bottom layer with a bottom image which is associated with this top image. When viewed through the transparent top layer both top and bottom images are seen as a composite image. At least one of the top and bottom layers include a plurality of image elements moveable relative to each other to produce at least one desired composite image which is the solution to the puzzle. For added complexity, both top and bottom layers may include moveable image elements which are manipulated from opposite sides of the puzzle. The puzzle is particularly adapted to slide puzzles having two layers of slide tiles.
Weimar: A puzzle is provided comprising a base formed from a transparent material and including a generally planar bottom wall and a frame extending upwardly from the bottom wall to define a puzzle recess. The puzzle further comprises a plurality of transparent puzzle pieces dimensioned to be received within the puzzle recess of the base. The transparent puzzle pieces are formed from a material that removably accepts markings thereon. The puzzle can be used by placing the base and the assembled puzzle pieces over a selected image which can be traced with a suitable marker. The puzzle pieces with the traced image thereon can be removed from the base and subsequently reassembled. The image created on the puzzle pieces can further be removed therefrom to enable the creation of a new puzzle.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RUPALI DHAKAD whose telephone number is (571)270-3743. The examiner can normally be reached M-F 8:30-5:30.
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, Alexander Lagor can be reached at 5712705143. 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.
/R.D./Examiner, Art Unit 2437
/ALI S ABYANEH/Primary Examiner, Art Unit 2437
Prosecution Insights