PARALLELIZED CONTEXT MODELLING USING INFORMATION SHARED BETWEEN PATCHES

Application No. 18/507,949

• Filed: Nov 13, 2023
• Priority: Jun 09, 2021 — continuation of PCTEP2021065394
• Examiner: SYED, FARHAN M
• Assignee: Huawei Technologies Co., Ltd.
• OA Round: 1 (Non-Final)

Office Action

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 .

Status of Claims
In response to communications filed on 13 November 2023, claims 1-24 are presently pending in the application, of which, claims 1, 10, 11, and 21-24 are presented in independent form.

Information Disclosure Statement
The information disclosure statement (IDS) submitted on 15 November 2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.

Drawings
The drawings, filed 13 November 2023, have been reviewed and accepted 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 Objections
Claims 10, 21, and 22 are objected to because of the following informalities: The claims appear to be independent claims and are written in shorthand to cover the features of the respective independent claims. For consistency purposes, the Examiner requires each and every feature to be in proper claim construction form. Appropriate correction is required.

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, 10, 11 and 21-24 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.
Claims 1, 10, and 22-23 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential steps, such omission amounting to a gap between the steps. See MPEP § 2172.01. The omitted steps are: Failure to establish what values are contained in a latent sensor prior to separating the latent sensor into a plurality of patches. Failure to identify or indicate the layers of neural network in how the elements are applied to each of the one or more layers of a neural network. Failure to identify how a probability model is developed and in the manner in which it is obtained for entropy encoding.

Claims 11, 21 and 24 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential steps, such omission amounting to a gap between the steps. See MPEP § 2172.01. The omitted steps are: Failure to establish what values are contained in a latent sensor prior to separating the latent sensor into a plurality of patches. Failure to identify or indicate the layers of neural network in how the elements are applied to each of the one or more layers of a neural network. Failure to identify how a probability model is developed and in the manner in which it is obtained for entropy encoding.

Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.

Claims 1-24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.

Regarding claims 1-9, under Step 2A claims 1-9 recite a judicial exception (abstract idea) that is not integrated into a practical application and does not provide significantly more.
Under Step 2A (prong 1), and taking claim 1 as representative, claim 1 recites:
separating the latent tensor into a plurality of patches each patch including one or more elements;
processing the plurality of patches by one or more layers of a neural network, including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel; and
obtaining a probability model for the entropy encoding of a current element of the latent tensor based on the processed set of elements.
These limitations recite mental processes, such as concepts performed in the human mind (see: 2019 PEG, p. 52). This is because the each of the limitations above recite a series of steps that may be mentally performed by which an evaluation is made for an abstract data. For example, the limitations of ‘separating the latent tensor into a plurality of patches each patch including one or more elements; processing the plurality of patches by one or more layers of a neural network, including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel; and obtaining a probability model for the entropy encoding of a current element of the latent tensor based on the processed set of elements,’ illustrate a judgement being performed to find matching results and does not perform any technical operation. This represents a judgement or decision which are concepts performed in the human mind and falls under certain methods of mental processes. Accordingly, under step 2A (prong 1) the claim recites an abstract idea because the claim recites limitations that fall within the “Certain methods of mental processes” grouping of abstract ideas (see again: 2019 PEG, p. 52).
Under Step 2A (prong 2), the abstract idea is not integrated into a practical application. The Examiner acknowledges that representative claim 1 does recite additional elements, including hardware processing circuitry, such as edge device.
Although reciting these additional elements, taken alone or in combination these elements are not sufficient to integrate the abstract idea into a practical application. This is because the additional elements of claim 1 are recited at a high level of generality (i.e. as generic computing hardware) such that they amount to nothing more than the mere instructions to implement or apply the abstract idea on generic computing hardware (or, merely uses a computer as a tool to perform an abstract idea). Further, the additional elements do no more than generally link the use of a judicial exception to a particular technological environment or field of use (such as the Internet or computing networks).
Secondly, the additional elements are insufficient to integrate the abstract idea into a practical application because the claim fails to (i) reflect an improvement in the functioning of a computer, or an improvement to other technology or technical field, (ii) implement the judicial exception with, or use the judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, (iii) effect a transformation or reduction of a particular article to a different state or thing, or (iv) applies or uses the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment.
In view of the above, under Step 2A (prong 2), claim 1 does not integrate the recited exception into a practical application (see again: 2019 Revised Patent Subject Matter Eligibility Guidance).
Under Step 2B, examiners should evaluate additional elements individually and in combination to determine whether they provide an inventive concept (i.e., whether the additional elements amount to significantly more than the exception itself). In this case, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. That is, the limitations of ‘obtaining a probability model for the entropy encoding of a current element of the latent tensor based on the processed set of elements,’ are additional elements that are insignificant extra solution activities that that do not amount to significantly more than the judicial exception.
Returning to representative claim 1, taken individually or as a whole the additional elements of claim 1 do not provide an inventive concept (i.e. they do not amount to “significantly more” than the exception itself). As discussed above with respect to the integration of the abstract idea into a practical application, the additional elements used to perform the claimed process amount to no more than the mere instructions to apply the exception using a generic computer and/or no more than a general link to a technological environment.
Furthermore, the additional elements fail to provide significantly more also because the claim simply appends well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception. For example, the additional elements of claim 1 utilize operations the courts have held to be well-understood, routine, and conventional (see: MPEP 2106.05(d)(lI)), including at least:
• receiving or transmitting data over a network, and/or
• storing and retrieving information in memory
• performing repetitive calculations
Even considered as an ordered combination (as a whole), the additional elements of claim 1 do not add anything further than when they are considered individually.
In view of the above, representative claim 1 does not provide an inventive concept (“significantly more”) under Step 2B, and is therefore ineligible for patenting.

Dependent claim 2 also does not integrate the abstract idea into a practical application. Notably, claim 2 recites ‘wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one; the set of elements has dimensions K x M corresponding to the K x M grid of patches and includes one element per patch within the subset of patches, and the one element is the current element; and the convolution kernel is a two-dimensional B x C convolution kernel, with B and C being positive integers, at least one of B and C being greater than one,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 2 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 2 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 2 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually.
In view of the above, claim 2 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 3 also does not integrate the abstract idea into a practical application. Notably, claim 3 recites ‘…wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one; the set of elements has dimensions L x K x M corresponding to the K x M grid of patches and includes L elements per patch including the current element and one or more previously encoded elements, wherein L is an integer greater than one; and the convolution kernel is three-dimensional A x B x C convolution kernel, with A being an integer greater than one and B and C being positive integers, at least one of B and C being greater than one, and wherein the method further comprises: storing the previously encoded elements in a historical memory’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 3 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 3 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 3 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually.
In view of the above, claim 3 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 4 also does not integrate the abstract idea into a practical application. Notably, claim 4 recites ‘reordering the elements included in the set of elements which are co-located in the subset of patches by projecting the co-located elements according to a respective patch position within the subset of patches onto a same spatial plane before the processing,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 4 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 8 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 4 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 4 do not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 5 also does not integrate the abstract idea into a practical application. Notably, claim 5 recites ‘wherein the convolution kernel is trainable within the neural network,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 5 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 5 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 5 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 5 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 6 also does not integrate the abstract idea into a practical application. Notably, claim 6 recites ‘ applying a masked convolution to each of the elements included in a patch out of the plurality of patches, which convolutes the current element and subsequent elements in the encoding order within the patch with zero,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 6 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 6 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 6 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 6 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 7 also does not integrate the abstract idea into a practical application. Notably, claim 7 recites ‘entropy encoding the current element into a first bitstream using the obtained probability model, and wherein the method further comprises: including the patch size into the first bitstream,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 7 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 7 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 7 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 7 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 8 also does not integrate the abstract idea into a practical application. Notably, claim 8 recites ‘wherein patches within the plurality of patches are non-overlapping and each patch out of the plurality of patches is of a same patch size, wherein the method further comprises: padding the latent tensor such that a new size of the latent tensor is a multiple of the same patch size before separating the latent tensor into the plurality of patches, and wherein the latent tensor is padded by zeroes,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 8 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 8 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 8 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 8 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 9 also does not integrate the abstract idea into a practical application. Notably, claim 9 recites ‘quantizing the latent tensor before separating into patches,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 9 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 9 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 9 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 9 do not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Claim 10 appear to include similar subject matter as in claim 1 as discussed above. More specifically, independent claim 10 additionally recites ‘a non-transitory machine-readable medium having instructions stored therein…’ which is recited at a high level of generality and are recited as performing mere generic computer functions routinely used in computer applications. Generic computer components recited as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system in addition to merely indicating a field of use or technological environment in which the judicial exception do not amount to significantly more than the exception itself. All the comments made with respect to the rejection of claim 1 equally apply and therefore stand rejected.

Claim 23 appear to include similar subject matter as in claim 1 as discussed above. More specifically, independent claim 23 additionally recites ‘an apparatus for entropy encoding for a latent tensor comprising …’ and is recited at a high level of generality and are recited as performing mere generic computer functions routinely used in computer applications. Generic computer components recited as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system in addition to merely indicating a field of use or technological environment in which the judicial exception do not amount to significantly more than the exception itself. All the comments made with respect to the rejection of claim 1 equally apply and therefore stand rejected.

Regarding claims 11-20, under Step 2A claims 11-20 recite a judicial exception (abstract idea) that is not integrated into a practical application and does not provide significantly more.
Under Step 2A (prong 1), and taking claim 11 as representative, claim 11 recites:
initializing the latent tensor with zeroes;
separating the latent tensor into a plurality of patches, each patch including one or more elements;
processing the plurality of patches by one or more layers of a neural network, including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel;
obtaining a probability model for the entropy decoding of a current element of the latent tensor based on the processed set of elements.
These limitations recite mental processes, such as concepts performed in the human mind (see: 2019 PEG, p. 52). This is because the each of the limitations above recite a series of steps that may be mentally performed by which an evaluation is made for an abstract data. For example, the limitations of ‘ initializing the latent tensor with zeroes; separating the latent tensor into a plurality of patches, each patch including one or more elements; processing the plurality of patches by one or more layers of a neural network, including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel,’ illustrate a judgement being performed to find matching results and does not perform any technical operation. This represents a judgement or decision which are concepts performed in the human mind and falls under certain methods of mental processes. Accordingly, under step 2A (prong 1) the claim recites an abstract idea because the claim recites limitations that fall within the “Certain methods of mental processes” grouping of abstract ideas (see again: 2019 PEG, p. 52).
Under Step 2A (prong 2), the abstract idea is not integrated into a practical application. The Examiner acknowledges that representative claim 11 does recite additional elements, including hardware processing circuitry, such as edge device.
Although reciting these additional elements, taken alone or in combination these elements are not sufficient to integrate the abstract idea into a practical application. This is because the additional elements of claim 11 are recited at a high level of generality (i.e. as generic computing hardware) such that they amount to nothing more than the mere instructions to implement or apply the abstract idea on generic computing hardware (or, merely uses a computer as a tool to perform an abstract idea). Further, the additional elements do no more than generally link the use of a judicial exception to a particular technological environment or field of use (such as the Internet or computing networks).
Secondly, the additional elements are insufficient to integrate the abstract idea into a practical application because the claim fails to (i) reflect an improvement in the functioning of a computer, or an improvement to other technology or technical field, (ii) implement the judicial exception with, or use the judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, (iii) effect a transformation or reduction of a particular article to a different state or thing, or (iv) applies or uses the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment.
In view of the above, under Step 2A (prong 2), claim 11 does not integrate the recited exception into a practical application (see again: 2019 Revised Patent Subject Matter Eligibility Guidance).
Under Step 2B, examiners should evaluate additional elements individually and in combination to determine whether they provide an inventive concept (i.e., whether the additional elements amount to significantly more than the exception itself). In this case, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. That is, the limitations of ‘obtaining a probability model for the entropy decoding of a current element of the latent tensor based on the processed set of elements,’ are additional elements that are insignificant extra solution activities that that do not amount to significantly more than the judicial exception.
Returning to representative claim 1, taken individually or as a whole the additional elements of claim 11 do not provide an inventive concept (i.e. they do not amount to “significantly more” than the exception itself). As discussed above with respect to the integration of the abstract idea into a practical application, the additional elements used to perform the claimed process amount to no more than the mere instructions to apply the exception using a generic computer and/or no more than a general link to a technological environment.
Furthermore, the additional elements fail to provide significantly more also because the claim simply appends well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception. For example, the additional elements of claim 11 utilize operations the courts have held to be well-understood, routine, and conventional (see: MPEP 2106.05(d)(lI)), including at least:
• receiving or transmitting data over a network, and/or
• storing and retrieving information in memory
• performing repetitive calculations
Even considered as an ordered combination (as a whole), the additional elements of claim 11 do not add anything further than when they are considered individually.
In view of the above, representative claim 11 does not provide an inventive concept (“significantly more”) under Step 2B, and is therefore ineligible for patenting.

Dependent claim 12 also does not integrate the abstract idea into a practical application. Notably, claim 12 recites ‘wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one; the set of elements has dimensions K x M corresponding to the K x M grid of patches and includes one element per patch within the subset of patches, and the one element is the current element; and the convolution kernel is a two-dimensional B x C convolution kernel, with B and C being positive integers, at least one of B and C being greater than one,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 12 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 12 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 12 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually.
In view of the above, claim 12 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 13 also does not integrate the abstract idea into a practical application. Notably, claim 13 recites ‘…wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one; the set of elements has dimensions L x K x M corresponding to the K x M grid of patches and includes L elements per patch including the current element and one or more previously encoded elements, wherein L is an integer greater than one; and the convolution kernel is three-dimensional A x B x C convolution kernel, with A being an integer greater than one and B and C being positive integers, at least one of B and C being greater than one; and wherein the method further comprises: storing of the previously decoded elements in a historical memory,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 13 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 13 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 13 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually.
In view of the above, claim 13 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 14 also does not integrate the abstract idea into a practical application. Notably, claim 14 recites ‘reordering the elements included in the set of elements which are co-located in the subset of patches by projecting the co-located elements according to a respective patch position within the subset of patches onto a same spatial plane before the processing,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 14 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 14 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 14 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 14 do not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 15 also does not integrate the abstract idea into a practical application. Notably, claim 15 recites ‘wherein the convolution kernel is trainable within the neural network,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 15 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 15 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 15 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 15 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 16 also does not integrate the abstract idea into a practical application. Notably, claim 16 recites ‘entropy decoding the current element from a first bitstream using the obtained probability model, and wherein the method further comprises: extracting the patch size from the first bitstream,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 16 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 16 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 16 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 16 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 17 also does not integrate the abstract idea into a practical application. Notably, claim 17 recites ‘wherein patches within the plurality of patches are non-overlapping and each patch out of the plurality of patches is of a same patch size, wherein the method further comprises: padding the latent tensor such that a new size of the latent tensor is a multiple of the same patch size before separating the latent tensor into the plurality of patches, and wherein the latent tensor is padded by zeroes,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 17 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 17 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 17 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 17 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 18 also does not integrate the abstract idea into a practical application. Notably, claim 18 recites ‘determining the probability model for the entropy decoding using: information of co-located elements that are currently to be decoded, or information of co-located current elements and information of co-located elements that have been previously decoded, and wherein the method further comprises: determining the probability model according to: information about previously decoded elements, and/or properties of the first bitstream,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 18 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 18 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 18 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 18 does not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 19 also does not integrate the abstract idea into a practical application. Notably, claim 19 recites ‘entropy decoding a hyper-latent tensor from a second bitstream; obtaining a hyper-decoder output by hyper-decoding the hyper-latent tensor; and wherein the method further comprises: separating the hyper-decoder output into a plurality of hyper-decoder output patches, each hyper-decoder output patch including one or more hyper-decoder output elements; concatenating a hyper-decoder output patch out of the plurality of hyper-decoder output patches and a corresponding patch out of the plurality of patches before the processing; and reordering the hyper-decoder output elements included in the set of elements which are co-located in the subset of patches by projecting the co-located elements onto the same spatial plane,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 19 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 19 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 19 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 19 do not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Dependent claim 20 also does not integrate the abstract idea into a practical application. Notably, claim 20 recites ‘wherein one or more of the following are performed in parallel for each patch out of the plurality of patches: applying the convolution kernel, and entropy decoding the current element,’ all which are more complexities descriptive of the abstract idea itself. Such complexities do not themselves provide further additional elements in addition to the abstract ideas themselves. Further, claim 20 relies upon at least similar additional elements that are mere instructions to implement the abstract idea or other exception on a computer. Considered both individually and as a whole, claim 20 does not integrate the recited exception into a practical application for at least similar reasons as discussed above.
Considered individually or as a whole, claim 20 also fail to result in “significantly more” than the abstract idea under step 2B. This is again because the claims merely recite additional elements that are insignificant extra-solution activity that apply the exception on generic computing hardware, generally link the exception to a technological environment, and append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception (see discussion above).
Even when viewed as an ordered combination (as a whole), the additional elements of the dependent claims do not add anything further than when they are considered individually. In view of the above, claim 20 do not provide an inventive concept (“significantly more”) under Step 2B, and are therefore ineligible for patenting.

Claim 21 appear to include similar subject matter as in claim 1 as discussed above. More specifically, independent claim 21 additionally recites ‘a non-transitory machine-readable medium having instructions stored therein…’ which is recited at a high level of generality and are recited as performing mere generic computer functions routinely used in computer applications. Generic computer components recited as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system in addition to merely indicating a field of use or technological environment in which the judicial exception do not amount to significantly more than the exception itself. All the comments made with respect to the rejection of claim 11 equally apply and therefore stand rejected.

Claim 24 appear to include similar subject matter as in claim 11 as discussed above. More specifically, independent claim 24 additionally recites ‘an apparatus for entropy encoding for a latent tensor comprising …’ and is recited at a high level of generality and are recited as performing mere generic computer functions routinely used in computer applications. Generic computer components recited as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system in addition to merely indicating a field of use or technological environment in which the judicial exception do not amount to significantly more than the exception itself. All the comments made with respect to the rejection of claim 11 equally apply and therefore stand rejected.

Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –

(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.

(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.

Claims 1-24 are rejected under 35 U.S.C. 102(a)(1)/(a)(2) as being unpatentable by Besenbruch, Chri, et al (U.S. 2022/0279183, filed May 10th, 2022, claiming priority to provisional application no. 63/017,295, filed 20 July 2020 and known hereinafter as Besenbruch).

As per claim 1, Besenbruch teaches a method for entropy encoding of a latent tensor, comprising:
separating the latent tensor into a plurality of patches each patch including one or more elements (e.g. Bensenbruch, see paragraph [0824-0829], which discloses an input that is provided by a neural network which encodes input image x and produces a latent representation y, where the quantized latent is entropy-encoded into a bitstream.);
processing the plurality of patches by one or more layers of a neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network.), including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel (e.g. Bensenbruch, see paragraphs [0855-0865], which discloses one or more layers of neural networks, where performance relates to file size, distortion, and runtime performance, where for example, see paragraph [0862], which discloses a layer of encoder neural network, the layer includes a convolution, a bias, and an activation function, where a subset of batches are factorized in a normal distribution.); and
obtaining a probability model for the entropy encoding of a current element of the latent tensor based on the processed set of elements (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

As per claim 2, Besenbruch teaches the method according to claim 1, wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one (e.g. Besenbruch, see paragraphs [0752-0757], which discloses an integer discrete flow transforming input x into z, split in z sub 1, z sub 2, and z sub 3.);
the set of elements has dimensions K x M corresponding to the K x M grid of patches and includes one element per patch within the subset of patches, and the one element is the current element (e.g. Besenbruch, see paragraphs [0788-0790], which discloses using convolution blocks that are displayed as grids and show cony-blocks as patches.); and
the convolution kernel is a two-dimensional B x C convolution kernel, with B and C being positive integers, at least one of B and C being greater than one (e.g. Besenbruch, see paragraphs [0811-0813], which discloses convolution kernels which encapsulates the composite convolution operations.).

As per claim 3, Besenbruch teaches the method according to claim 1, wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one (e.g. Besenbruch, see paragraphs [0752-0757], which discloses an integer discrete flow transforming input x into z, split in z sub 1, z sub 2, and z sub 3.);
the set of elements has dimensions L x K x M corresponding to the K x M grid of patches and includes L elements per patch including the current element and one or more previously encoded elements, wherein L is an integer greater than one (e.g. Besenbruch, see paragraphs [0788-0790], which discloses using convolution blocks that are displayed as grids and show cony-blocks as patches.); and
the convolution kernel is three-dimensional A x B x C convolution kernel, with A being an integer greater than one and B and C being positive integers, at least one of B and C being greater than one (e.g. Besenbruch, see paragraphs [0811-0813], which discloses convolution kernels which encapsulates the composite convolution operations.), and wherein the method further comprises:
storing the previously encoded elements in a historical memory (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

As per claim 4, Besenbruch teaches the method according to claim 1, further comprising:
reordering the elements included in the set of elements which are co-located in the subset of patches by projecting the co-located elements according to a respective patch position within the subset of patches onto a same spatial plane before the processing (e.g. Bensenbruch, see paragraphs [0754-0757], which discloses shuffling the order of the channels in the feature map.).

As per claim 5, Besenbruch teaches the method according to claim 1, wherein the convolution kernel is trainable within the neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network which can be trained.).

As per claim 6, Besenbruch teaches the method according to claim 1, further comprising:
applying a masked convolution to each of the elements included in a patch out of the plurality of patches, which convolutes the current element and subsequent elements in the encoding order within the patch with zero (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 7, Besenbruch teaches the method according to claim 1, further comprising:
entropy encoding the current element into a first bitstream using the obtained probability model (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.), and
wherein the method further comprises: including the patch size into the first bitstream (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 8, Besenbruch teaches the method according to claim 1, wherein patches within the plurality of patches are non-overlapping and each patch out of the plurality of patches is of a same patch size, wherein the method further comprises:
padding the latent tensor such that a new size of the latent tensor is a multiple of the same patch size before separating the latent tensor into the plurality of patches, and wherein the latent tensor is padded by zeroes (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 9, Besenbruch teaches the method according to claim 1, further comprising:
quantizing the latent tensor before separating into patches (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 10, Besenbruch teaches a method for encoding image data comprising:
obtaining a latent tensor by processing the image data with an autoencoding convolutional neural network (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.); and
entropy encoding the latent tensor into a bitstream using a generated probability model according to claim 1 (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 11, Besenbruch teaches a method for entropy decoding of a latent tensor, comprising:
initializing the latent tensor with zeroes (e.g. Bensenbruch, see paragraph [0773], which discloses when decoded, the laten space is denormalized using the same mean and variance, where the laten space is normalized and trained against a normal prior distribution with mean zero and variance 1.);
separating the latent tensor into a plurality of patches, each patch including one or more elements (e.g. Bensenbruch, see paragraph [0824-0829], which discloses an input that is provided by a neural network which encodes input image x and produces a latent representation y, where the quantized latent is entropy-encoded into a bitstream.);
processing the plurality of patches by one or more layers of a neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network.), including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel (e.g. Bensenbruch, see paragraphs [0855-0865], which discloses one or more layers of neural networks, where performance relates to file size, distortion, and runtime performance, where for example, see paragraph [0862], which discloses a layer of encoder neural network, the layer includes a convolution, a bias, and an activation function, where a subset of batches are factorized in a normal distribution.);
obtaining a probability model for the entropy decoding of a current element of the latent tensor based on the processed set of elements (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

As per claim 12, Besenbruch teaches the method according to claim 11, (e.g. Besenbruch, see paragraphs [0752-0757], which discloses an integer discrete flow transforming input x into z, split in z sub 1, z sub 2, and z sub 3.);
the set of elements has dimensions K x M corresponding to the K x M grid of patches and includes one element per patch within the subset of patches, and the one element is the current element (e.g. Besenbruch, see paragraphs [0788-0790], which discloses using convolution blocks that are displayed as grids and show cony-blocks as patches.); and
the convolution kernel is a two-dimensional B x C convolution kernel, with B and C being positive integers, at least one of B and C being greater than one (e.g. Besenbruch, see paragraphs [0811-0813], which discloses convolution kernels which encapsulates the composite convolution operations.).

As per claim 13, Besenbruch teaches the method according to claim 11, wherein the subset of patches forms a K x M grid of patches, with K and M being positive integers, at least one of K and M being greater than one (e.g. Besenbruch, see paragraphs [0752-0757], which discloses an integer discrete flow transforming input x into z, split in z sub 1, z sub 2, and z sub 3.);
the set of elements has dimensions L x K x M corresponding to the K x M grid of patches and includes L elements per patch including the current element and one or more previously encoded elements, wherein L is an integer greater than one (e.g. Besenbruch, see paragraphs [0788-0790], which discloses using convolution blocks that are displayed as grids and show cony-blocks as patches.); and
the convolution kernel is three-dimensional A x B x C convolution kernel, with A being an integer greater than one and B and C being positive integers, at least one of B and C being greater than one (e.g. Besenbruch, see paragraphs [0811-0813], which discloses convolution kernels which encapsulates the composite convolution operations.), and wherein the method further comprises:
storing the previously decoded elements in a historical memory (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

As per claim 14, Besenbruch teaches the method according to claim 11, further comprising:
reordering the elements included in the set of elements which are co-located in the subset of patches by projecting the co-located elements according to a respective patch position within the subset of patches onto a same spatial plane before the processing (e.g. Bensenbruch, see paragraphs [0754-0757], which discloses shuffling the order of the channels in the feature map.).

As per claim 15, Besenbruch teaches the method according to claim 11, wherein the convolution kernel is trainable within the neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network which can be trained.).

As per claim 16, Besenbruch teaches the method according to claim 11, further comprising:
entropy decoding the current element from a first bitstream using the obtained probability model (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.), and wherein the method further comprises: extracting the patch size from the first bitstream (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 17, Besenbruch teaches the method according to claim 11, wherein patches within the plurality of patches are non-overlapping and each patch out of the plurality of patches is of a same patch size, wherein the method further comprises:
padding the latent tensor such that a new size of the latent tensor is a multiple of the same patch size before separating the latent tensor into the plurality of patches, and wherein the latent tensor is padded by zeroes (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 18, Besenbruch teaches the method according to claim 16, further comprising: determining the probability model for the entropy decoding using:
information of co-located elements that are currently to be decoded, or information of co-located current elements and information of co-located elements that have been previously decoded (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.), and
wherein the method further comprises: determining the probability model according to:
information about previously decoded elements, and/or properties of the first bitstream (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 19, Besenbruch teaches the method according to claim 11, further comprising:
entropy decoding a hyper-latent tensor from a second bitstream (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.);
obtaining a hyper-decoder output by hyper-decoding the hyper-latent tensor (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.); and
wherein the method further comprises:
separating the hyper-decoder output into a plurality of hyper-decoder output patches, each hyper-decoder output patch including one or more hyper-decoder output elements (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.);
concatenating a hyper-decoder output patch out of the plurality of hyper-decoder output patches and a corresponding patch out of the plurality of patches before the processing (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.); and
reordering the hyper-decoder output elements included in the set of elements which are co-located in the subset of patches by projecting the co-located elements onto the same spatial plane (e.g. Bensenbruch, see paragraphs [0754-0757], which discloses shuffling the order of the channels in the feature map.).

As per claim 20, Besenbruch teaches the method according to claim 11, wherein one or more of the following are performed in parallel for each patch out of the plurality of patches:
applying the convolution kernel, and entropy decoding the current element (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 21, Besenbruch teaches a method for decoding image data comprising:
entropy decoding a latent tensor from a bitstream according to claim 11 (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.); and
obtaining the image data by processing the latent tensor with an autodecoding convolutional neural network (e.g. Bensenbruch, see paragraphs [0631-0639], which discloses encoding the input image using a first trained neural network to produce a late representation, and to identify nonlinear convolution kernals, entropy encoding the quantized latent and an identification of the identified nonlinear convolution kernals into a bitstream; transmitting the bitstream to produce the quantized latent.).

As per claim 22, Besenbruch teaches a non-transitory computer readable medium having code instructions stored thereon, which, upon execution by one or more processors, cause the one or more processors to execute the method separating the latent tensor into a plurality of patches each patch including one or more elements (e.g. Bensenbruch, see paragraph [0824-0829], which discloses an input that is provided by a neural network which encodes input image x and produces a latent representation y, where the quantized latent is entropy-encoded into a bitstream.);
processing the plurality of patches by one or more layers of a neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network.), including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel (e.g. Bensenbruch, see paragraphs [0855-0865], which discloses one or more layers of neural networks, where performance relates to file size, distortion, and runtime performance, where for example, see paragraph [0862], which discloses a layer of encoder neural network, the layer includes a convolution, a bias, and an activation function, where a subset of batches are factorized in a normal distribution.); and
obtaining a probability model for the entropy encoding of a current element of the latent tensor based on the processed set of elements (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

As per claim 23, Besenbruch teaches an apparatus for entropy encoding of a latent tensor, comprising:
a processing circuitry configured to:
separating the latent tensor into a plurality of patches each patch including one or more elements (e.g. Bensenbruch, see paragraph [0824-0829], which discloses an input that is provided by a neural network which encodes input image x and produces a latent representation y, where the quantized latent is entropy-encoded into a bitstream.);
processing the plurality of patches by one or more layers of a neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network.), including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel (e.g. Bensenbruch, see paragraphs [0855-0865], which discloses one or more layers of neural networks, where performance relates to file size, distortion, and runtime performance, where for example, see paragraph [0862], which discloses a layer of encoder neural network, the layer includes a convolution, a bias, and an activation function, where a subset of batches are factorized in a normal distribution.); and
obtaining a probability model for the entropy encoding of a current element of the latent tensor based on the processed set of elements (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

As per claim 24, Besenbruch teaches an apparatus for entropy decoding of a latent tensor, comprising:
a processing circuitry configured to:
initializing the latent tensor with zeroes (e.g. Bensenbruch, see paragraph [0773], which discloses when decoded, the laten space is denormalized using the same mean and variance, where the laten space is normalized and trained against a normal prior distribution with mean zero and variance 1.);
separating the latent tensor into a plurality of patches, each patch including one or more elements (e.g. Bensenbruch, see paragraph [0824-0829], which discloses an input that is provided by a neural network which encodes input image x and produces a latent representation y, where the quantized latent is entropy-encoded into a bitstream.);
processing the plurality of patches by one or more layers of a neural network (e.g. Bensenbruch, see paragraphs [0825-0830], which discloses one or more layers of neural network.), including processing a set of elements which are co-located in a subset of patches by applying a convolution kernel (e.g. Bensenbruch, see paragraphs [0855-0865], which discloses one or more layers of neural networks, where performance relates to file size, distortion, and runtime performance, where for example, see paragraph [0862], which discloses a layer of encoder neural network, the layer includes a convolution, a bias, and an activation function, where a subset of batches are factorized in a normal distribution.);
obtaining a probability model for the entropy decoding of a current element of the latent tensor based on the processed set of elements (e.g. Bensenbruch, see paragraph [0798], which discloses obtaining a probability model using an encoder and decoder, which transforms the quantized latent into a bitstream. See further paragraph [0792], which discloses a visualized tensor perspective.).

Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure. See attached PTO-892 that includes additional prior art of record describing the general state of the art in which the invention is directed to.

Contact Information
Any inquiry concerning this communication or earlier communications from the examiner should be directed to FARHAN M SYED whose telephone number is [phone]. The examiner can normally be reached M-F 8:30AM-5:30PM.
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, Apu Mofiz can be reached at 571-272-4080. 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.

/FARHAN M SYED/Primary Examiner, Art Unit 2161 June 14, 2026

Prosecution Timeline

• Nov 13, 2023: Application Filed
• Jun 22, 2026: Non-Final Rejection mailed — §101, §102, §112 (current)