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 .
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55 (Chinese PCT Application CN2022/106139 filed July 16th, 2022).
Response to Arguments
Applicant canceled claim 20.
Applicant added new claim 21.
The pending claims are 1 – 19 and 21 [Page 7 lines 1 – 6].
Applicant amended the Title of the Invention to address Examiner’s Specification Objection [Page 7 lines 7 – 12].
Applicant's arguments filed April 16th, 2026 [Page 7 lines 13 – Page 8 line 9] have been fully considered but they are not persuasive.
The Applicant contends the “conversion” is clear, however, the term to encompass encoding and decoding (in dependent claims 16 and 17) are readily recognized as inverse processes to one of ordinary skill in the art (see Specification Paragraph 57) contrary to the Applicant’s assertion of parallel processes [Page 7 lines 17 – 25]. Thus using inverse processes create inconsistencies as no closed definitions are provided in the Specification thus in view of MPEP2173.03 [“In addition, inconsistencies in the meaning of terms or phrases between claims may render the scope of the claims to be uncertain. Tvngo Ltd. (BVI) v. LG Elecs. Inc., 861 Fed. Appx. 453, 459-60, 2021 USPQ2d 697 (Fed. Cir. 2021) ("The issue is not breadth of the dependent claims but their use of the disputed phrase in a way that contradicts the independent claims.”]. Regarding the second point of the Applicant, the Examiner refers to the fact the processes are inverses and in view of MPEP2173.03, finds claim 1 Indefinite [Page 8 lines 1 – 9].
While the Applicant’s points maybe understood, the Examiner respectfully disagrees; thus the Rejection has been maintained.
Applicant’s arguments, see Page 8 line 10 – Page 9 line 12, filed April 16th, 2026, with respect to claim 1 have been fully considered and are persuasive. The 35 USC 112(b) Essential Steps Rejection of claims 1 – 20 has been withdrawn.
The argument is not necessarily persuasive for the Applicant’s arguments, but rather in view of MPEP 2172.01 the Specification hardly recognizes critical or essential elements [“Rejections under 35 U.S.C. 112(b) (or pre-AIA 35 U.S.C. 112, second paragraph) for failing to claim the subject matter that the inventor or a joint inventor regards as the invention based on the omission of essential matter must explain the basis for concluding that the inventor regards the omitted matter to be essential to the invention. See, e.g., Ex parte Robertson, Appeal 2016-001938, op. at 4 (PTAB 2017)(Rejection reversed because examiner failed to identify where in the specification the unclaimed elements were described as essential. Board noted that summary of the invention in the specification did not include features alleged by examiner to be essential.)”].
Applicant’s arguments, see Page 9 line 13 – Page 10 line 12, filed April 16th, 2026, with respect to claim 1 have been fully considered and are persuasive. The 35 USC 112(b) Indefinite Rejection of claim 18 has been withdrawn.
While the argument is deemed persuasive, the Examiner observes in the First and Third points [Page 9 lines 13 – 30 and Page 10 lines 16 – 20], the Applicant argues the limitation is an Intended use recitation which will not receive patentable weight. In view of MPEP2181 IA [“When applicant uses the term "means" or "step" in the preamble, a rejection under 35 U.S.C. 112(b) may be appropriate when it is unclear whether the preamble is reciting a means- (or step-) plus- function limitation or whether the preamble is merely stating the intended use of the claimed invention. When applicant merely states an intended use of the claimed invention in the preamble (e.g., "A device for printing, comprising ..."), the examiner should not construe such language as reciting a means-plus-function limitation.”], the Examiner in the sole interest to expedite prosecution affords no patentable weight to the preamble as reciting an intended use.
The Examiner notes in the arguments [Page 10 lines 1 – 15], the Applicant appears to agree with the Examiner’s analysis the claimed “processor” does NOT invoke Functional Analysis under 112(f) thus in the interest of brevity, the section is removed.
Applicant's arguments filed April 16th, 2026 [Page 10 line 27 – Page 13 line 19] have been fully considered but they are not persuasive.
Examiner’s Notes: Reproduced Figures and their captions are not included in the line numberings provided.
The Applicant in canceling claim 20 renders Examiner’s 102 Rejection moot.
First, the Applicant cites the references against the claims [Page 10 lines 21 – 26] and then without Specification support cites the disputed limitation of claim 18 [Page 10 lines 27 – 30].
Second, the Applicant cites portions of Jia Page 52 and Figures 17 – 18 as not rendering obvious the “target neighboring samples being in the same region” feature in claim 18 [Page 11 lines 1 – Page 12 line 11]. The Examiners notes the reproduced sections are Jia are similar to the teachings of the Applicant’s Specification in Figures 19 – 20 (see at least reference characters 2010 and sub-characters) in which samples across region boundaries are used as described in Specification Paragraphs 54 – 56. Additionally when considering Jia as a whole Pages 49 – 50 teach the samples in L (the overlapped region) as used to compute those in R (the smaller non-overlapping regions) as shown in cited Jia Figures 14 – 16 and 19. Considering Jia as a whole, Figures 14 – 16 and 19 which were also cited in the Rejection of the disputed feature render obvious non-overlapping regions and partitions of an image into regions / tiles for processing similar to Applicant’s teachings in Figures 15 – 17 (subfigures included) in which Jia in at least cited Pages 32 – 33 (tile based partitioning in encoding / decoding) and Pages 49 – 50 which are related to the target samples and region determinations are made and the regions are non-overlapping as required in claim 18. Additionally in dependent claim 3, overlapping regions or using samples from different regions of the current sample is permitted and thus within scope of claim 1, thus Jia as a whole renders obvious claims 1 and 3.
Third, the Applicant argues against Alshina’s padding operations and teachings [Page 12 lines 12 – 18]. However, the Examiner disagrees as Jia already teaches the disputed features claimed and further the embodiments of the Applicant’s Specification rely on padding as taught by Jia and Alshina as well and further considering Alshina as a whole Paragraphs 185 – 186 render obvious the disputed features of the regions being non-overlapping for processing.
Fourth, the Applicant contends Jia and Alshina do not render obvious the disputed feature of claim 18 [Page 12 line 19 – Page 13 line 2], recites new claim 21 to replace canceled claim 20 [Page 13 lines 3 – 16], and contends all claims are allowable [Page 13 lines 17 – 19]. The Examiner disagrees for at least the reasons given above.
Applicant's arguments do not comply with 37 CFR 1.111(c) because they do not clearly point out the patentable novelty which he or she thinks the claims present in view of the state of the art disclosed by the references cited (Ikonin noted argued) or the objections made. Further, they do not show how the amendments avoid such references (e.g. Ikonin) or objections.
While the Applicant’s points maybe understood, the Examiner respectfully disagrees; thus the Rejection has been maintained.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on January 15th, 2025 was filed before the mailing date of the First Action on the Merits (mailed January 16th, 2026). The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the Examiner.
Specification
The lengthy specification has not been checked to the extent necessary to determine the presence of all possible minor errors. Applicant’s cooperation is requested in correcting any errors of which applicant may become aware in the specification.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1 – 15, 18 – 19, and 21 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 1, the claim recites "conversion" which has Indefinite metes and bounds as processes (e.g. encoding) and their inverse (e.g. decoding) are encompassed thus the claim has Indefinite metes and bounds as the steps of the claim have no distinction between such processes.
Regarding claims 18 – 19 and 21, see claim 1 which performs the steps of the claimed apparatus (claim 18), program (claim 19), and method (claim 21) and thus are similarly Rejected.
Regarding claims 2 – 15, the dependent claims do not cure the deficiencies of their respective independent claims and thus are similarly Rejected.
While claims 16 and 17 are not Rejected, the inclusion of only one of the respective dependent claims (not both simultaneously) would overcome the Rejection of claim 1 and the recitation of the other would result in improper dependency issues (e.g. a decoder depending on an encoder as the inverse process depending on the forward process or vice versa).
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 (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
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 – 19 and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Jia, et al. (WO2024/002497 A1 referred to as “Jia” throughout) [First cited in the Office Action mailed January 16th, 2026], and further in view of Ikonin, et al. (US PG PUB 2023/0262243 A1 referred to as “Ikonin” throughout) and Alshina, et al (US PG PUB 2023/0353766 A1 referred to as “Alshina” throughout).
Regarding claim 1, see claim 18 which is the apparatus performing the steps of the claimed method.
Regarding claim 19, see claim 18 which is the apparatus performing the steps of the claimed program.
Regarding claim 21, see claim 18 which is the apparatus performing the steps of the claimed method steps or is similarly rejected to claim 1 as the method claimed where the bitstream storing / generating is taught by Jia Figures 1 – 3 and 23 – 24 (see at least reference characters 2400, 2410, and 2416) as well as Page 15 line 4 – Page 16 line 11 (processors for encoding / decoding with memory storing instructions), Page 62 line 23 – Page 63 line 17 (system with programming to generate or store a bitstream) and Page 99 line 32 – Page 100 line 30 (non-transitory medium storing a program for execution to generate / read / store a bitstream) as the non-transitory medium storing a program to perform a method to generate a bitstream.
Regarding claim 18, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
a processor [Jia Figures 1 – 4 (see at least reference characters 230, 302, and 2410) as well as Page 15 line 21 – Page 16 line 11 (processors for encoding / decoding)] and
a non-transitory memory with instructions thereon, wherein the instructions upon execution by the processor [Jia Figures 1 – 3 and 23 – 24 (see at least reference characters 260, 304, 2400, 2410, and 2416) as well as Page 15 line 4 – Page 16 line 11 (processors for encoding / decoding with memory storing instructions), Page 62 line 23 – Page 63 line 17 (system with programming to generate a bitstream) and Page 99 line 32 – Page 100 line 30 (non-transitory medium storing a program for execution to generate / read a bitstream)], cause the processor to perform acts comprising:
obtaining, for a conversion between visual data and a bitstream of the visual data [Jia Figures 1 – 6 (subfigures included and see at least reference characters 20 and 30), 9 (see at least reference characters 901, 902, 904, 905, and 906), and 21 – 25 as well as Page 13 line 22 – Page 14 line 19 (encoder / decoder), Page 18 line 3 – Page 19 line 10 (encoding / decoding picture (obvious variant of images see Page 1 lines 15 – 32) and video as the claimed visual data), Page 31 line 31 – Page 32 line 25 (auto-en(de)coder / en(de)coder network), and Page 36 line 18 – Page 37 line 2 (decoder / encoder as NN / NN in an encoding / decoding framework)], region information indicating sizes of a plurality of regions in a quantized latent representation of the visual data [Jia Figures 15 – 20 (partitions of times) and 25 – 26 as well as Pages 68 – 76 (tables in Pages 70 – 76 are included which include information regarding the size (width and height) of tiles and Pages 68 – 69 describe partitioning images / latents into tiles and tile size considerations) and 77 – 79 (see Page 77 lines 23 – 29 (size using width and height) and Page 79 line 1 – 18 (see at least the “image_tile.size” syntax element)];
selecting, based on the region information, a set of target neighboring samples from a plurality of candidate neighboring samples of a current sample in the quantized latent representation [Jia Figures 10 – 16 and 17 – 21 (subfigures included) as well as Page 37 lines 7 – 21 (padding (target samples) based on current samples in the latent for NN processing), 51 (based on region sizes / determination of regions sizes for NN processing where Pages 49 – 50 provide additional region information considerations in view of at least Figures 15 – 16 and 19), Page 52 lines 8 – 35 (target samples available including nearest neighbor in the same region / tile); Alshina Figures 16 – 17, 21, and 23 as well as Paragraphs 292 – 303 (see at least reflection padding techniques which uses samples in the same region / tile)],
the set of target neighboring samples being in the same region as the current sample [Jia Figures 10 – 14, 15 – 16 (cited as part of the region information with the current sample) and 17 – 21 (subfigures included) as well as Page 37 lines 7 – 21 (padding (target samples) based on current samples in the latent for NN processing), 51 (based on region sizes (further described in Pages 49 – 50) / determination of regions sizes for NN processing (which includes the use of non-overlapping regions as in Pages 49 – 50 and Figures 15 – 16 and 19 as cited for the selecting based on region limitation for the current and target / neighboring samples)), Page 52 lines 8 – 35 (target samples available including nearest neighbor in the same region / tile); Alshina Figures 16 – 17, 21, and 23 as well as Paragraphs 292 – 303 (see at least reflection padding techniques which uses samples in the same region / tile to process latents – combinable with Jia)];
determining the current sample based on the set of target neighboring samples [Jia Figures 14 – 21 (subfigures included – see bitstream1 and bittream2 generated at least) as well as Page 32 line 18 – Page 33 line 24 (samples processed in quantized latent representation and context modelling present), Page 37 lines 7 – 21 (padding (target samples) based on current samples in the latent for NN processing), 51 (based on region sizes / determination of regions sizes for NN processing where non-overlapped regions and expansion are taught in Pages 49 – 50 (Figures 15 – 16 and 19)), Page 52 lines 8 – 35 (target samples available including nearest neighbor in the same region / tile), and Page 67 lines 9 – 30 (padding to reconstruct current samples based on target neighbor samples); Alshina Figures 16 – 17, 21, and 23 as well as Paragraphs 292 – 303 (see at least reflection padding techniques which uses samples in the same region / tile – to combine with Jia); Ikonin Figures 3 and 13 as well as Paragraphs 135 – 139 (probability modelling for the entropy coding / decoding to determine the current sample uses and auto-regressive process) and 188 – 192 (autoregressive modelling / processing used in a decoder half / affects encoding the current sample)]; and
performing the conversion based on the current sample [See Jia citations above regarding “conversion” and additionally Jia Figure 9 as well as Page 32 line 18 – Page 33 line 24 (samples processed in quantized latent representation and context modelling present) in combination with Ikonin Figures 3 and 13 as well as Paragraphs 135 – 139 (probability / context modelling for the entropy coding / decoding to determine the current sample uses and auto-regressive process) and 188 – 192 (autoregressive modelling / processing used in a decoder half / affects the context model)].
The motivation to combine Ikonin with Jia is to combine features in the same / related field of invention of neural network processing and side information for video compression [Ikonin Paragraphs 3 – 5] in order to improve compression using neural networks [Ikonin Paragraphs 7 – 8 and 107 where the Examiner observes KSR Rationales (D) or (F) are also applicable].
The motivation to combine Alshina with Ikonin and Jia is to combine features in the same / related field of invention of neural networks for image encoding / decoding [Alshina Paragraphs 2 – 4] in order to improve efficiency in NN processing [Alshina Paragraphs 7 – 10 where the Examiner observes KSR Rationales (D) or (F) are also applicable].
This is the motivation to combine Jia, Ikonin, and Alshina which will be used throughout the Rejection.
Regarding claim 2, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the current sample is a quantized latent sample of the visual data [Jia Figure 9 as well as Page 30 line 20 – Page 31 line 24 (samples of quantized latent currently processed in encoder / decoder) and Page 32 line 26 – Page 33 line 24 (samples processed in quantized latent representation)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 3, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the determination of the current sample and a determination of a further sample in the quantized latent representation is allowed to be performed in parallel [Jia Figures 15 – 21 as well as Page 41 lines 4 – 26 (wavefront processing to process regions / tiles in parallel in further combination / view of Ikonin Paragraph 154), Page 45 lines 13 – 35 (parallel processing regions / tiles and samples therein), Page 58 lines 17 – 35 (parallel processing of tiles / samples in latents), and Page 66 lines 1 – 31 (tile independently coded / decoded and done in parallel)], and
the further sample is located in a region different from a region in which the current sample located [].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 4, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the current sample is determined by using an auto-regressive process [Jia Figure 9 as well as Page 32 line 18 – Page 33 line 24 (samples processed in quantized latent representation and context modelling present) in combination with Ikonin Figures 3 and 13 as well as Paragraphs 135 – 139 (probability modelling for the entropy coding / decoding to determine the current sample uses and auto-regressive process) and 188 – 192 (autoregressive modelling / processing used in a decoder half / affects encoding the current sample)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 5, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the auto-regressive process is a context model or a multistage context model [Jia Figure 9 as well as Page 32 line 18 – Page 33 line 24 (samples processed in quantized latent representation and context modelling present) in combination with Ikonin Figures 3 and 13 as well as Paragraphs 135 – 139 (probability / context modelling for the entropy coding / decoding to determine the current sample uses and auto-regressive process) and 188 – 192 (autoregressive modelling / processing used in a decoder half / affects the context model)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 6, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the region information is determined [See claim 1 or 18 “region information” limitation for citations] based on at least one of the following:
a depth of a transform that is performed to obtain a latent representation of the visual data [Jia Page 41 lines 4 – 18 (levels of wavefront processing) and Page 59 line 7 – Page 60 line 4 or alternatively Ikonin Paragraphs 114 – 116 (depth of kernels / filters to perform transforms)],
the number of regions in the plurality of regions [Jia Page 90 lines 14 – 30],
the sizes of the plurality of regions [Jia Figures 14 – 18 (subfigures included) as well as Page 84 line 1 – Page 85 line 8 (size / height / width signaled and used in a tile / region size determination)],
positions of the plurality of regions [Jia Figures 14 – 18 (subfigures included) as well as Page 84 line 1 – Page 85 line 8 (start positions of the region / tile used for size determinations)],
a size of the latent representation [Jia Figures 6 – 11 as well as Page 35 line 12 – Page 36 line 17],
a size of the quantized latent representation [Jia Figures 6 – 11 as well as Page 35 line 12 – Page 36 line 17],
a size of a reconstruction of the visual data [Jia Figures 6 – 11 as well as Page 35 line 12 – Page 36 line 17],
a color format of the visual data [Jia Page 17 lines 28 – 34],
a color component of the visual data [Jia Page 59 line 7 – Page 60 line 4 and Page 97 line 30 – Page 99 line 31], or
information regarding whether the visual data is resized [Jia Page 89 line 20 – Page 90 line 3 and Page 93 line 25 – Page 94 line 19 (signaling resizing ratio); Alshina Paragraphs 287 – 294 (resizing determinations and signaling)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 7, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the bitstream comprises at least one indication associated with the number of regions in the plurality of regions [Jia Page 81 line 27 – Page 82 line 29 (signaling the number of tiles / regions)], or
wherein the bitstream comprises at least one indication associated with the sizes of the plurality of regions [Jia Figures 14 – 18 (subfigures included) as well as Page 84 line 1 – Page 85 line 8 (size / height / width signaled and used in a tile / region size determination)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 8, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
determining statistical information of the current sample based on the set of target neighboring samples [Jia Figures 6, 9, and 16 as well as Page 34 lines 1 – 23 (computing statistical properties / information including means and variances), Page 55 line 23 – Page 56 line 5 (combine with the use of target samples in Page 52 lines 21 – 25 (target samples in the latent / partition)), and Page 66 line 8 – Page 67 line 14 (encoding / decoding the current sample based on the statistical information computed); Ikonin Figures 12 – 13 as well as Paragraphs 187 – 192 (code segment included in which the side information / statistics are used to determine current samples)]; and
determining the current sample based on the statistical information [Jia Figures 6, 9, 11, 16 – 20, and 23 – 24 as well as Page 52 lines 21 – 25 (target samples in the latent / partition) and Page 66 line 8 – Page 67 line 14 (encoding / decoding the current sample based on the statistical information computed); Ikonin Figures 12 – 13 as well as Paragraphs 187 – 192 (code segment included in which the side information / statistics are used to determine current samples); Alshina Figures 4 – 6 as well as Paragraphs 186 and 213 – 221 (mean and variance computed as statistical properties used to compute samples)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 9, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the plurality of candidate neighboring samples are dependent on a processing kernel used to processing the current sample [Jia Figures 6 – 9, 11, and 16 – 19 (subfigures included) as well as Page 49 line 22 – Page 51 line 12 (crop / pad latent to fit the processing kernel size / dimensions to combine with Page 52 line 21 – Page 53 line 22)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 10, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
determining values for a part of samples in the processing kernel based on values for the set of target neighboring samples [Jia Figures 6 – 9, 11, and 16 – 19 (subfigures included) as well as Page 49 line 22 – Page 51 line 12 (crop / pad latent to fit the processing kernel size / dimensions to combine with Page 52 line 21 – Page 53 line 22 (target samples in the latent / partition))];
determining values for the rest of the samples in the processing kernel based on a predetermined value [Jia Figures 6 – 9, 11, and 16 – 19 (subfigures included) as well as Page 49 line 22 – Page 51 line 12 (crop / pad latent to fit the processing kernel size / dimensions to combine with Page 52 line 21 – Page 53 line 22 (constant / zero fill values)); Alshina Paragraphs 51 – 55, 92 – 96, and 255 – 259 (filling / padding with 0 values where one of ordinary skill understands zero as a constant / fixed term)]; and
determining the statistical information based on values for the samples in the processing kernel [Jia Figures 6, 9, and 16 as well as Page 34 lines 1 – 23 (computing statistical properties / information including means and variances), Page 55 line 23 – Page 56 line 5 (combine with the use of target samples in Page 52 lines 21 – 25 (target samples in the latent / partition)), and Page 66 line 8 – Page 67 line 14 (encoding / decoding the current sample based on the statistical information computed); Ikonin Figures 12 – 13 as well as Paragraphs 187 – 192 (code segment included in which the side information / statistics are used to determine current samples and the autoregressive approach /model based on NN / kernels)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 11, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the predetermined value is constant [Jia Figure 11 as well as Page 52 lines 21 – 35 (constant / zero fill values); Alshina Paragraphs 51 – 55, 92 – 96, and 255 – 259 (filling / padding with 0 values where one of ordinary skill understands zero as a constant / fixed term)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 12, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the predetermined value is 0 [Jia Figure 11 as well as Page 52 lines 21 – 35 (constant or zero fill values); Alshina Paragraphs 51 – 55, 92 – 96, and 255 – 259 (filling / padding with 0 values)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 13, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the transform comprises one of the following:
an analysis transform [Ikonin Figures 1 – 3 as well as Paragraphs 135 – 137 (analysis / synthesis transform)],
a wavelet-based forward transform [Jia Figure 15 as well as Page 41 lines 4 – 26 (wavefront processing an obvious variant of the claimed wavelet transform) or alternatively Ikonin Figures 1 – 3 as well as Paragraph 132 (suggests using a wavelet transform)], or
a discrete cosine transform (DCT) [Jia Page 29 line 25 – Page 30 line 3 (DCT suggested to use as a transform); Ikonin Figures 5 – 7 and 21 as well as Paragraphs 142 – 144, 150 (using DCT transforms) and 250 – 257 (transforms to use including DCTs)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 14, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the statistical information [See claim 1 for citations] comprises at least one of the following:
a mean value [See next limitation for citations], or a variance [Jia Figures 6 (subfigures included) and 9 as well as Page 33 lines 1 – 24 (statistical side information of quantized latents computed included means or variances) and Page 60 lines 5 – 21 (means and variances computed and affects entropy tables / contexts used); Alshina Figures 4 – 6 as well as Paragraphs 186 and 219 – 221 (mean and variance computed as statistical properties)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 15, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the visual data comprise a picture of a video or an image [Jia Figures 4 – 5 (see at least reference characters 20 and 30) as well as Page 18 line 3 – Page 19 line 10 (encoding / decoding picture (obvious variant of images see Page 1 lines 15 – 32) and video)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 16, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the conversion includes encoding the visual data into the bitstream [Jia Figures 1 – 6 (subfigures included and see at least reference character 20) and 9 (see at least reference characters 901, 902, 904, 905, and 906) as well as Page 13 line 22 – Page 14 line 19 (encoder), Page 31 line 31 – Page 32 line 25 (auto-encoder / encoder network), and Page 36 line 18 – Page 37 line 2 (encoder as NN / NN in an encoding framework)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Regarding claim 17, Jia teaches a neural network (NN) based coding / decoding approach using quantized latents and pads the regions / tiles partitions and computes statistics affecting the context model / coding method used. Ikonin teaches additional considerations in using padding and the use of auto-regressive processed in NNs. Alshina teaches signaling considerations on the size of regions / tiles for processing.
It would have been obvious to one of ordinary skill art before the effective filing date of the claimed invention to modify the teachings of Jia to include auto-regressive configurations as taught by Ikonin and to signal partitioning information as taught by Alshina. The combination teaches
wherein the conversion includes decoding the visual data from the bitstream [Jia Figures 1 – 6 (subfigures included and see at least reference character 30) and 9 (see at least reference characters 901, 902, 904, 905, and 906) as well as Page 13 line 22 – Page 14 line 19 (decoder), Page 31 line 31 – Page 32 line 25 (auto-decoder / decoder network), and Page 36 line 18 – Page 37 line 2 (decoder as NN / NN in a decoding framework)].
See claim 1 for the motivation to combine Jia, Ikonin, and Alshina.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Reference considered which may raise ODP issues based on amendments to the claims include: Esenlik, et al. (US PG PUB 2024/0430482 A1 referred to as “Esenlik” throughout) and Esenlik, et al. (US PG PUB 2024/0430428 A1 referred to as “Esen” throughout).
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Tyler W Sullivan whose telephone number is (571)270-5684. The examiner can normally be reached IFP.
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, David Czekaj can be reached at (571)-272-7327. 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.
/TYLER W. SULLIVAN/ Primary Examiner, Art Unit 2487
Prosecution Insights