www.uspto.gov
`
`UNITED STATES DEPARTMENT OF COMMERCE
`United States Patent and TrademarkOffice
`Address; COMMISSIONER FOR PATENTS
`P.O. Box 1450
`Alexandria, Virginia 22313-1450
`
`16/664,084
`
`10/25/2019
`
`Takahiro NISHI
`
`2019-1804A
`
`7418
`
`CP
`Lind&
`Wenderoth,
`Wenderoth, Lind & Ponack, L.L.P.
`1025 Connecticut Avenue, NW
`Suite 500
`Washington, DC 20036
`
`KALAPODAS, DRAMOS
`
`2487
`
`PAPER NUMBER
`
`NOTIFICATION DATE
`
`DELIVERY MODE
`
`03/03/2021
`
`ELECTRONIC
`
`Please find below and/or attached an Office communication concerning this application or proceeding.
`
`The time period for reply, if any, is set in the attached communication.
`
`Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
`following e-mail address(es):
`eoa@ wenderoth.com
`kmiller@wenderoth.com
`
`PTOL-90A (Rev. 04/07)
`
`

`

`
`
`Disposition of Claims*
`1,3-12 and 14-24 is/are pending in the application.
`)
`Claim(s)
`5a) Of the above claim(s) ___ is/are withdrawn from consideration.
`C] Claim(s)__ is/are allowed.
`Claim(s) 1,3-12 and 14-24 is/are rejected.
`(1 Claim(s)__is/are objectedto.
`C} Claim(s)
`are subjectto restriction and/or election requirement
`* If any claims have been determined allowable, you maybeeligible to benefit from the Patent Prosecution Highway program at a
`participating intellectual property office for the corresponding application. For more information, please see
`http://www.uspto.gov/patents/init_events/pph/index.jsp or send an inquiry to PPHfeedback@uspto.gov.
`
`) ) ) )
`
`Application Papers
`10)() The specification is objected to by the Examiner.
`11) The drawing(s) filed on 10/25/2019 is/are: a)[¥) accepted or b)() objected to by the Examiner.
`Applicant may not request that any objection to the drawing(s) be held in abeyance. See 37 CFR 1.85(a).
`Replacement drawing sheet(s) including the correction is required if the drawing(s) is objected to. See 37 CFR 1.121 (d).
`
`Priority under 35 U.S.C. § 119
`12)Z) Acknowledgment is made of a claim for foreign priority under 35 U.S.C. § 119(a)-(d) or (f).
`Certified copies:
`_—_c)L) None ofthe:
`b)L) Some**
`a)X) All
`1.2 Certified copies of the priority documents have been received.
`2.2) Certified copies of the priority documents have been received in Application No.
`3.2.) Copies of the certified copies of the priority documents have been receivedin this National Stage
`application from the International Bureau (PCT Rule 17.2(a)).
`* See the attached detailed Office action for a list of the certified copies not received.
`
`Attachment(s)
`
`1)
`
`Notice of References Cited (PTO-892)
`
`2) (J Information Disclosure Statement(s) (PTO/SB/08a and/or PTO/SB/08b)
`Paper No(s)/Mail Date
`U.S. Patent and Trademark Office
`
`3) (J Interview Summary (PTO-413)
`Paper No(s)/Mail Date
`(Qj Other:
`
`4)
`
`PTOL-326 (Rev. 11-13)
`
`Office Action Summary
`
`Part of Paper No./Mail Date 20210226
`
`Application No.
`Applicant(s)
`16/664,084
`NISHI etal.
`
`Office Action Summary Art Unit|AIA (FITF) StatusExaminer
`DRAMOS KALAPODAS
`2487
`Yes
`
`
`
`-- The MAILING DATEofthis communication appears on the cover sheet with the correspondence address --
`Period for Reply
`
`A SHORTENED STATUTORY PERIOD FOR REPLYIS SET TO EXPIRE 3 MONTHS FROM THE MAILING
`DATE OF THIS COMMUNICATION.
`Extensions of time may be available underthe provisions of 37 CFR 1.136(a). In no event, however, may a reply betimely filed after SIX (6) MONTHSfrom the mailing
`date of this communication.
`If NO period for reply is specified above, the maximum statutory period will apply and will expire SIX (6) MONTHSfrom the mailing date of this communication.
`-
`- Failure to reply within the set or extended period for reply will, by statute, cause the application to become ABANDONED (35 U.S.C. § 133}.
`Any reply received by the Office later than three months after the mailing date of this communication, evenif timely filed, may reduce any earned patent term
`adjustment. See 37 CFR 1.704(b).
`
`Status
`
`1) Responsive to communication(s) filed on 12/18/2020.
`C} A declaration(s)/affidavit(s) under 37 CFR 1.130(b) was/werefiled on
`2a)¥) This action is FINAL.
`2b) (J This action is non-final.
`3)02 An election was madeby the applicant in responseto a restriction requirement set forth during the interview
`on
`; the restriction requirement and election have been incorporated into this action.
`4\0) Since this application is in condition for allowance except for formal matters, prosecution as to the merits is
`closed in accordance with the practice under Exparte Quayle, 1935 C.D. 11, 453 O.G. 213.
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 2
`
`DETAILED ACTION
`
`Notice of Pre-AlA or AIA Status
`
`1.
`
`The present application, filed on or after March 16, 2013, is being examined
`
`under the first inventor to file provisions of the AIA.
`
`Claim Status
`
`2.
`
`Claims 1, 3-12 and 14-24 are currently pending
`
`Claims 2 and 13 have been cancelled.
`
`The rejection to claims 7 and 18 under 35 U.S.C. 112(b) is withdrawn upon
`
`amendment
`
`Responseto Arguments
`
`3.
`
`Applicant’s arguments with respect to the rejection(s) of claims 1, 3-12 and 14-24
`
`have been fully considered and are unpersuasive.
`
`3-1. Applicants’ Argument
`
`i
`
`Applicants allege thal “Hippel discloses technology of improving the accuracy
`
`achieved by an encoder and a discriminator by a GAN for image encoding technology.
`
`Although, Fa. 3 of Rippel Hustrates that Prediction 318 is output fram Discriminator
`
`34, Applicant notes that Prediction 318 indicates whether the input to Discriminator
`
`304 is an original image or a reconstructed image (See [Q006]}. Accordingly, His
`
`respectiuily submitted that Prediction 218 is nel a predicted image.”
`
`Ep.
`
`Therefore, Applicant notes that Discriminator 304 of Rippel discriminates
`
`whether the input is an original image or a reconstructed image, and as such, if
`
`is respectiully submitted that Rippel falis to teach that Discriminator 304 outputs 3
`
`probability that a predicted image, which is nether an original image nor a
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 3
`
`reconstructed image, matches an original image to feed back the probability to
`
`the generator network.
`
`Applicant respectfully submits that any combination of Terada and Rinpe!
`
`would, at best, teach encoding an original image using the image encading
`
`apparatus of Terada, generating a reconstructed image by further decoding the
`
`encoded ariginal image, and adjusting neural networks of predictors (.¢., the NN
`
`intra predictor and the NN inter predictor of Terada} of the image encoding
`
`apparatus, based on a result of discriminating whether the input to the discrirninator
`
`is an original image or a reconstructed image as taught by Rippel.
`
`i}
`
`However, itis respectiully submitted that any combination of Terada and
`
`Hippel faiis io teach thal a predicted image is generated using a generator network,
`
`and that a probability that the predicted image maiches an original image (e.q., the
`
`inpul image required by claim 1} is fed back ic the generator nebwork using a
`
`discriminator network in order te update the generator network.
`
`Accordingly, any combination of Terada and Hippel necessarily fails to teach “feeds
`
`back, io the generator network, a probability that the predicted image matches
`
`the inpul unage Oy inputting the inopul image and the predicted image to a
`
`discriminator network, the discriminator network being a neural network and
`
`constituting a generative adversarial network (GAN) with the generator network" and
`
`“updaies the generaior nebvark and the discriminator network lo reduce difference
`
`between the inoul image and the predicted image and increase accuracy af
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 4
`
`discriminating between the inoul image and the predicted image,” as required by the
`
`above-noted features of claim 1.
`
`3-2.
`
`Examiner’s Rebuttal
`
`To point 3-1(i)
`
`To this point, Examiner contends that the arguments raised by Applicants to the
`
`mapping of claim 2, as being regarded by the Office to have referenced the paragraphs
`
`(e.g., Par.[0006]) or the figures (¢.g., Fig.3) or the discriminator 304 and its output 318
`
`herein allegedly introduced, are directed in error to the respective references captured
`
`from Rippel by which advancing the rationale that the Office would have improperly
`
`interpret Rippel, in considering the output (Id. 318) of the discriminator (Id. 304)
`
`represent a predicted image in a different embodiment seen asirrelevant by the Office.
`
`The argued paragraphs and figures are not part of the First Action on merit
`
`hence the responseto the Office Action of 07/21/2020, and at this point the argumentis
`
`considered mootfor failing to address the subject matter referenced in the 35 U.S.C.
`
`103 rejection.
`
`Please refer to MPEP 707.07 Unpersuasive Argument: Applicant Obtains
`Result Not Contemplated by Prior Art
`In responseto applicant's argumentthat [the recognized advantages over
`paragraphsnot used or referenced by the Office in rejection as stated at point 3-
`1(i) and repeatedat point 3-1(ii)], the fact that applicant has recognized another
`advantage which would flow naturally from following the suggestion of the prior art
`cannot be the basis for patentability when the differences would otherwise be obvious.
`See Ex parte Obiaya, 227 USPQ 58, 60 (Bd. Pat. App. & Inter. 1985).
`
`To point 3-1(ii)
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 5
`
`in rebuttal, ibis remarked that in rejecting claim 2 herein addressed by
`
`amenament, Examiner did not reference/mapped the claimed matter analysis to the
`
`discriminator element 304, nor misinterpreted the probability error output by unit 304
`
`as being a predicted image as alleged herein, hence the presumptive logic
`
`introduced by Applicants in argumentciting; “any combination of Terada and Rippel
`
`necessarily fails to teach.....”, may not be considered relevant from its referencing a
`
`different embodimentof the art in responseto the Office action rejection to claim 2.
`
`See the evidence mappedat the respective claim.
`
`Thus, - Applicant's arguments fail to comply with 37 CFR 1.111(b) because they
`
`amount to a general allegation that the claims define a patentable invention without
`
`specifically pointing out how the combination art distinguishes from the claim.
`
`Furthermore, the Applicant's specific analysis of the art to Rippel, being directed
`
`to different portions identified from the respective description, have not been considered
`
`in the Office Action (O.A.) rejection. See rebuttal at points 3-1(i) and (ii), addressing the
`
`combined references of Terada and Rippel, while the argument fails to provide an
`
`evidenciary based responseto the rejection albeit the consel legal argument provided to
`
`the 35 U.S.C. 103 rejection.
`
`Please refer to; MPEP 700, (707.07(f) Answer All Material Traversed
`Unpersuasive Argument: No Teaching, Suggestion, or Motivation To Combine
`In responseto applicant's argumentthat there is no teaching, suggestion, or motivation
`to combine the references, the examiner recognizes that obviousness may be
`established by combining or modifying the teachings of the prior art to produce the
`claimed invention where there is some teaching, suggestion, or motivation to do so
`foundeither in the references themselves or in the knowledge generally available to one
`of ordinary skill in the art. See /n re Fine, 837 F.2d 1071, 5 USPQ2d 1596(Fed. Cir.
`1988), In re Jones, 958 F.2d 347, 21 USPQ2d 1941 (Fed. Cir. 1992), and KSR
`International Co. v. Teleflex, Inc., 550 U.S. 398, 82 USPQ2d 1385 (2007). In this
`case, [In support to the combinedarts to Terada and Rippel, it is found that based
`on suggestionsin Rippel the art to Terada teaches the matter claimed about
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 6
`
`applying the input image and the predicted image to the NN prediction parameter
`determiner 109 in Fig.1 and outputs the parameter to theintra/inter predictor of
`the encoder 300 in Fig.1, and Fig. 35 which is being feed-backedto the intra/inter
`NN predictors 110 and 111 respectively, Par.[0243]-[0246]].
`
`To point 3-1(iii)
`
`Examiner reiterates the evidentiary probe mappedat claim 2 presently
`
`embeddedinto claim 1 by amendment, and remarks that the claim rajection has been
`
`defended solely based on counsel argument corresponding to the claim language.
`
`The argument is directed to the provisions in MPEP 2145stipulating; “However,
`
`arguments of counsel cannottake the place of factually supported objective evidence.
`
`See, e.g., In re Huang, 100 F.3d 135, 139-40, 40 USPQ2d 1685, 1689 (Fed. Cir.
`
`1996); In re De Blauwe, 736 F.2d 699, 705, 222 USPQ 191, 196 (Fed. Cir. 1984).”
`
`Conclusion to Finality decision
`
`In lieu of the above rebuttals to the Applicant’s arguments referring to other
`
`portions submitting a new portion of the reference to argue that the combined arts to
`
`Terada and Rippel would in fact teach away asrelated to the refernce used by Office,
`
`Examiner constructs the finality of the rejection based on supplementing the original
`
`rejection based on responding to the argument provided.
`
`Applicant's representative is encouraged to contact the Exmainer with matter
`
`deemedto advance the prosecution.
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 7
`
`Claim Rejections - 35 USC § 103
`
`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.
`
`The following is a quotation of 35 U.S.C. 103 which forms the basis for all
`
`obviousnessrejections 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 ordinaryskill in
`the art to which the claimed invention pertains. Patentability shall not be negated by the manner
`in which the invention was made.
`
`The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148
`
`USPQ 459 (1966), that are applied for establishing a background for determining
`
`obviousness under 35 U.S.C. 103 are summarized asfollows:
`
`1. Determining the scope and contents of the prior art.
`2. Ascertaining the differences between the prior art and the claims at issue.
`3. Resolving the level of ordinary skill in the pertinent art.
`4. Considering objective evidence presentin the application indicating
`obviousness or nonobviousness.
`
`4.
`
`Claims 1, 3-12 and 14-24 are rejected under 35 U.S.C. 103 as being
`
`unpatentable over Kengo Teradaetal., (hereinafter Terada) (US 2018/0184123)in
`
`view of Oren Rippeletal., (hereinafter Rippel) (US 2018/0174052) in lieu of Prov.
`
`App. 62/434,600, 62/434,602, 62/434,603 and 62/458,749.
`
`Re Claim 1. (Currently Amended) Terada discloses, an encoder, comprising (an
`
`encoder 105 in Fig.35 or encoder 10 Fig.69A):
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 8
`
`orocessing circully fan enceding processor 100 in Fig.i}; and
`
`memory, wherein using ihe memory, the processing circultry Grame memory
`
`ti2 in Fig.t and step 5165 in Fig.6 Par.[O728}):
`
`generates a predicted image of an input image that is a current image to be
`
`encoded (generating an intra/inter predicted image, steps S121, Si22 Fig.3 by the
`
`intra/inter prediction generator Par(0119] of the current image to be encoded at
`
`aoparatus 100 in Fig.2 where the predicted image is generated by the NN in intra
`
`mode per Fig.4, of inter made in Fig.5}, based on generated data outoul from a
`
`generaior nelwork (and being based on the data outpul from the mode network
`
`generators for intra/inter prediction al Fig. 4 and 5 respectively, the encoding
`
`being based on a neural network, NN ¢.g., as generated by a function # outputting
`
`the prediction mode, per Fig.7 Par [00136] according to a parameter determiner
`
`109 setting different prediction and coding NN modes, Par [O1127) in response ta a
`
`reference image being input to the generator network, the generator nefwork being a
`
`neural network (the image prediction being based on the cutpul data generated by
`
`the neural network, NN generator in response to 4 reference image extracted at
`
`Sis, Fig.6, and inputted to the NN at Fig.13, Par.{Gi34] and on the program code
`
`ai Fig.19 Par.[O1S2] to generate the predicted pixel ¢.g., in “nan_inira’ mode and
`
`Fig.?4 Par{0145] and cade at Fig.20 for generating the prediction image in
`
`“aninfer’ mode being neural networks, Par [0153] and/or [O143h:
`
`calculates a prediction error by subtracting the predicted image from the input
`
`image (a prediction error is computed by subiraciing the predicied image from the
`
`inpul current image, Par {O104], (0105) [0120]124), (O1S6), (O1Se8i, Fig.7): andl
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 9
`
`generaies an encoded image by at least transiorming the prediction error (and
`
`further generating the encoded image at 105, Fig.i1, Par.[0136], based on
`
`prediction differences Fig.7, Par [0104] at encoder 100 generating the image
`
`stream af 105 by applying frequency transform at 103, step S726, on the
`
`prediction error Le., being computed at the diference block S125, at subtractor
`
`102 wherein generating the prediction error is generated by subtracting the
`
`prediction block from the current block, Par {O1G4] [0120
`
`feeds back, to the generator network, a probability that the predicted image
`
`maiches the indut image by inputting the input image and the predicted imaue fo a
`
`discriminator network Geeding back to the intra/inter NN generator the feedback
`
`parameters from NN by matching the inpul image to the predicted image Fig.7
`
`START
`
`
`
`
`INPUT CURRENTBLOCK AND NEIGHBORINGPIXELS
`TO|
`NEURAL NETWORK, AND DETERMINEPARAMETERFOR|
`“ CALCULATINGPE
`‘Bt
`NEIGHBOR
`
`
`
`“END ___
`A
`
`
`Par (2124), [O1S0-[01 32] or [0736], or based on highest correlation in case of inter
`
`NN orediction, Par O139]-(0141p.
`
`wodaies the generator network and the discriminator network to reduce
`
`difierence between ihe input image and the predicted image and increase accuracy of
`
`discriminating bebveen the input image and the predicted image (updating the
`
`generator network “NWN intra/inter Predictor’ 110 and f11 and the discriminator
`
`network 1694 with data from the input image and data fram the predicted image
`
`i.e., the prediction block, in Fig.35 to increase the accuracy of the predicted image
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 10
`
`by reducing the difference error, or discriminating between the input anc
`
`predicted image io be small per step S172 in fig.7 herein reproduced with
`
`highlights far brevity
`
`FIG. 36
`
`r3
`
`DIFFERENCE
`“ean
`BLOCK PAR
`;,
`SUBTRAGCTOR o
`Dna
`
`Le HOSA
`
`pi BITRTREAM
`
`and Par [0243)-[0246)).
`
`Within the same neural network prediction and coding apparatus and method
`
`based on Neural Network error feedback being determined belweeri thie input and the
`
`oredicied image described in Terada, the art io Rippel expressly teaches about using a
`
`generative adversarial network (GAN} with the generator network feeding back, to the
`
`generator network a probability that the predicted image matches the input image by
`
`inputting the input image and the predicted wage to a discriminator nebwark,
`foeds back, oredicted wmace to the generator network, 4 orobabiliy that the
`
`
`
`
`
`matches the inout image by inputting the input image and the predicted image to a
`
`discriminator network
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 11
`
`the discriminator network being a neural network and constituting a qenerative
`
`
`
`adversarialnetwork(GAN)withthegeneratornetwork (@ generative adversarial
`
`network, Title, Abstract): and
`
`Wodates the generator network and the discriminator network to reduce
`
`difisrence between ihe input image and the predicted image and increase accuracy of
`
`predicted unace (updating the
`ciscrimmating bebveen the in Ui image and the
`
`
`
`
`discriminaior and the encoder, io minimize the loss, Par fO014] Le., fo reduce the
`
`difference between the criginal image and the reconstructed content, as earlier
`
`taught at Par.[O010] per discriminators 704 and the feedback output 734 in Fig.7
`tt Fa
`AMutoencederA t
`ctOS
`
` Saenannnannvngnenannanannene
`
`
`ies =’ Me PSR
`
`
`\ Aukoanooder
`TOR
`
`ft
`f
`
`‘ i
`
`
`
`vue
`renee
`
`
`
`. Par[0073] in order
`“EURAY
`to increase the prediction accuracy by reducing the error corresponding with the
`
`encoder loss, Par (0075-0077).
`
`in carsideration to ihe prediction process using neural networks identified in
`
`Terada being apmlied to reduce the prediction error (Par iGi3ep in order to improve the
`
`coding efficiency (Par [0158)) by switching the encoding modes for the best image
`
`quality (Par [0445] (0446) Fig. 22} and respectively reducing the image distortion,
`
`indicating the presence of the error feedback cetermined af the NN prediction 109A,
`
`and feedback fo the NN inter/intra predictors 110 and 111, e.g., being viewed as the NN
`
`orediclan discriminator, by which the ardinary skilled in the art would have had the
`
`incentive before the effective filing date of the application, io search for similar NN
`
`prediction modes seeking to reduce the difference between the current image and the
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 12
`
`uredicied value as identified in Rippel disclosing a NN coding method and reducing the
`
`prediction error by using a discriminator error feedback (Par.[0076],[0077]) thus
`
`deeming the combination predictable, hence obviating the claim.
`
`The rationale fo combine finds support in the Graham factual inquines necessary
`
`io substantiate the above combination, in view of the instant fact case under
`
`consideration and in accordance with explaining the conclusion of obviousness in view
`
`of the provisions stioulated in MPEP 2143: Basic Requirements of a Prima Facie Case
`
`of Obviousness. I.
`
`EXEMPLARY RATIONALES(A),(D) and (G),that may support a
`
`conclusion of obviousness above evicenced,
`
`including :
`
`(A) Combining prior art elements according to knawn methads to yield
`&
`predictable results improving the prediction error processing by applying
`feedback mechanism to the NN prediction method as disclosed in Rippel
`(Par.[0076],[0077]) by combining with Terada suggestions for error reduction
`{Par (0445) 10446] Fig 22};
`
`{(D) Applying @ known technique to a known device (method, or product}
`&
`reacy for improvernent to yield predictable results Gwhere the method in Rippel is
`known and adapted in the art per the preliminary provisional teachings}:
`
`(G) Some teaching, suqgestion, or motivation in the orior art that would
`®
`have led one of arcinary skill fo modify the prior art reference or ia combine prior art
`reference teachings ta arrive at the claimed invention (where the suggestion to
`combine relies on the common interest to reduce the prediction error identified in
`both aris}.
`
`See precedence im “The Federal Circuit recognized Agrizap as “a textbook case of when
`the asserted claims involve a cornbination of familiar elements according to known methads
`that does no more than yield predictable results.” id. Agrizap exemplifies a strang case of
`obviousness based on simple substitution that was not overcome by the objective evidence of
`nonobviousness offered. It also demonstrates that analogous art is not limited to the field of
`applicant’s endeavor, in that one of the references that usec an animal body as a resistive
`switch ta complete a circuit for the generation of an electric charge was not in the field of pest
`
`control”
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 13
`
`2, (Cancelled)
`
`Re Claim S. (Original Terada anc Alppel disclose, the encocler according to
`
`claim 1. wherein
`
`Terada teaches about, the reference image is a processed image included in a
`
`picture, the picture including the input image, anc
`
`in generating the predicied image, ihe processing circultry generates a first inira-
`
`precicied image as the predicted image, based an ihe generated data (ihe
`
`neighboring pixels represent ihe processed reference image included in the
`
`picture that includes the input image by which the first intra-predicted data is
`
`generated, Fig.7 and 14 per block 170 in Fig.1 Par.f0108)-[0709)] ¢.¢., as indicated
`
`by ihe intra-prediction switch for “Fixed intra- Predictor” L.e., regular intra-mode
`
`at block 116b in Fig.t Par [O121)-[01 23).
`
`Re Claim 4. (Original Terada and Rippel disclose, the encoder according to
`
`claim 3, wherein the processing circultry further:
`
`Terada ieaches about the processor, generales a second intra-precicted image
`
`of the input image by intra prediction based on the reference image (the processor
`
`generates the second intra-predicied image by switching from 110b te the 11Ga,
`
`NA prediction mode, Par [fO124)):
`
`selects an image from among the first inira-precicted image and the second
`
`inira- predicied image (selecting one of the plurality of indicated intra-predictecd
`
`inpul inages Fig.22, ¢.g., a second predicted image block, by applying the
`
`instruction code fer fixed “intra_pred_type” based on the neighboring references
`
`ai 1106, or indicating which one of the NN intra-prediction is used, Par jO154]
`
`according to the accuracy level determined at Par.[0i24] as depicied at switching
`
`block 110a in Fig.t and Par.[ot24)}: and
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 14
`
`when the processing circuitry selects the second imtra-predicted image in
`
`selecting the image, calculates the prediction errar by sublracting the second intra-
`
`predicied image from the input image in calculating the prediction error (he selected
`
`second intra-predicied block a1 110 in Fig.i, 11a, Le., the second intra-predicted
`
`image block, encodes the prediction error according to the NN, Par.{0120}).
`
`Re Claim 5. (Orginal Terada and Riopel disclose, the encoder according to
`
`clairn 3, wherein
`
`Terada teaches wherein, in generating the precictec image, the processing
`
`circuilry generates, as the first intra~ predicted image, the generated data cutout from
`
`ihe generaior nelwark in response to the reference image being inpul to the generaior
`
`network (processing Le., generating the first selected inira-predicted block Le., the
`
`first intre-predicted image block, encedes the prediction error, in regard te the
`
`selecied reference image inputted to the NN generator, Par [0120] Fig.13}.
`
`Re Claim 6. (Original) Terada and Rippel disclose, the encoder according ta
`
`claim 3, wherein
`
`Terada teaches wherein, in generating the predicted image, the processing
`
`clroultry: oblains, as an intra prediction parameter, the generated data outout from the
`
`generator network in response to the reference image being input to the generator
`
`network; and generates the first intra-predicied image by intra prediction based on the
`
`reference image and the intra prediction parameter (see Fig.7, obtaining Le.,
`
`determining ine intra-prediction parameter, based on which generating the first
`
`inira-predicied imaqe, Far.[O1i4i and Par fi30Liidi] and Fig.13 Par [O134p.
`
`Re Glairn 7. (Currently Amended} Terada and Rippel disclose, the encoder
`
`according fo claim 1, wherein
`
`Terada teaches abaul, the reference image is included in a processed picture
`
`thal is not etterertivem a picture that includes the input image, the processed picture
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 15
`
`being 2 picture on which the encoder has already performed processing related io
`
`encoding. ard
`
`in generating the precicied image, the processing circullry generates a first inter-
`
`predicied image as the predicted image, based on the generated data Gn an inter-
`
`prediction mode at block 111 relies on a reference picture applying the encoding
`
`process at units 103-104 then inte the prediction loop at 107-106 as reference
`
`picture appled to the NN inter-predictor tita, in Fig.1 and Par [O116)-[(0120) by
`
`which generating a first predicted image in inter-prediction mode based on the
`
`generated reference image data, Par (O137]-(0739]}.
`
`Re Claim 8. (Original Terada and Rippel disclose, the encoder accarding ta
`
`claim 7, wherein the processing circuttry further:
`
`Terada teaches about the processor, generates a second inter-precicted image
`
`of the input image by inter prediction based an the reference image (generating a
`
`second prediction by switching between “fixed inter-precdiction” mode P11 in
`
`Fig.?, where generating a second predicted image in inter-prediction mode based
`
`on the reference image per Par [0137]-(6139] according to condHions set al
`
`Par (O17sh:
`
`selects an image fram among the first inter-predicied image and the second
`
`infter- predicted image (selecting form first and second inter-prediction modes titb
`
`and 1i1a, or using the second layer to which a reference pixel is connected to a
`
`node is used as a second-inter-predicted pixel, Par [O178). and
`
`when the processing circultry selecis the second inter-predicted image in
`
`selecting the image, calculates the prediction error oy subtracting the second inter-
`
`predicted image from the input image in calculating the prediction error {upon selecting
`
`the second inter-predicted image, perform the calculation of the prediction errar
`
`as previously established at Par.[0170) (However, this claim is aise rejected
`
`under 35 U.S.C.112(3) for depending from claim 7 and requires clarification prior
`
`io assessing a proper examination).
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 16
`
`Re Claim 9. (Orginal Terada anc Rippel disclose, the encocler according to
`
`clair 7, wherein
`
`Terada icaches about, in generating the predicied image, ihe processing
`
`circuilry generates, as the first inter-~ predicted image, the generated data cutout from
`
`ihe generaior nelwark in response to the reference image being inpul to the generaior
`
`network (he infer-prediction mode selected al the NN position Tita in Fig.d block
`
`ti and is disclosed at Par fO11S], [0120] and executed per NN generator block in
`
`Fig.i4 Par (6143), (0145).
`
`Re Claim 10. (Orginal Terada and Rippel disclose, the encocer according to
`
`claim 7, wherein in generating the predicted image, the nrocessing circuitry:
`
`Terada teaches about, oblains, as an inter prediction parameter (per syntax of
`
`parameter in NN inter-prediction code at Fig.1S and 20), the generated data output
`
`from the gerieraior network in response to the reference image being input fo the
`
`generator network (obtaining the inter-prediciion parameter, the data generated at
`
`the ouiput of the NN according to the inputted image, Fig.14, Par [0943)}: and
`
`generates the first inter-predicted image by inter prediction based on the
`
`reierence image and ihe infer prediction’ parameter (generating the first inter-
`
`predicted image based on the reference and the NN inter-prediction parameter,
`
`Par (O137E(0139i and Fig.8}.
`
`Re Claim 11. (Orginal Terada and Rippel disclose, the encoder according to
`
`claim 1, wherein
`
`Terada teaches about, the generator network is a hierarchical network that
`
`includes an inpul layer, a hidden layer, and an outpul layer (ihe NN generator is of a
`
`hierarchical type, per Fig.d3, 14}.
`
`

`

`Application/Control Number: 16/664,084
`Art Unit: 2487
`
`Page 17
`
`Re Claim 12. (Currently Amended) This claim represents the decacting pari of the
`
`prediction loop made part of the encoder represented in claim 1, hence perfarming a
`
`similar prediction orocess in the decoding of the reconstructed picture by flowing the
`
`same Hmiting steps and in similar orderthus it is relected on the same evidentiary
`
`premise, mutatis mutandis.
`
`13. (Cancelled)
`
`Re Claim 14. (Original) This claim represents the decoding part of the prediction
`
`loop made part of the encoder represented in a imitation of claim 3, hence performing a
`
`similar prediction process in the decoding of the reconstructed picture by flowing the
`
`same limiting steps and in similar order thus it is relected on the same evidentiary
`
`premise, mutatis mutandis.
`
`Re Claim 15. (Original) This claim represenis the decoding part of the prediction
`
`loop mace part of the encoder represented in clairn 4, hence performing a similar
`
`brediclion process in the decoding of the reconstructed picture by Towing ihe same
`
`limiting stens and in similar orcer thus it is rejected on the same evidentiary premise,
`
`mutatis mutandis.
`
`Re Claim 16. (Original This c